Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

A systematic review of real-world diabetes prevention programs: learnings from the last 15years

A systematic review of real-world diabetes prevention programs: learnings from the last 15years Background: The evidence base for the prevention of type 2 diabetes mellitus (T2DM) has progressed rapidly from efficacy trials to real-world translational studies and practical implementation trials over the last 15 years. However, evidence for the effective implementation and translation of diabetes programs and their population impact needs to be established in ways that are different from measuring program effectiveness. We report the findings of a systematic review that focuses on identifying the critical success factors for implementing diabetes prevention programs in real-world settings. Methods: A systematic review of programs aimed at diabetes prevention was undertaken in order to evaluate their outcomes using the penetration, implementation, participation, and effectiveness (PIPE) impact metric. A search for relevant articles was carried out using PubMed (March 2015) and Web of Science, MEDLINE, CENTRAL, and EMBASE. A quality coding system was developed and included studies were rated independently by three researchers. Results: Thirty eight studies were included in the review. Almost all (92 %) provided details on participation; however, only 18 % reported the coverage of their target population (penetration). Program intensity or implementation—as measured by frequency of contacts during first year and intervention duration—was identified in all of the reported studies, and 84 % of the studies also reported implementation fidelity; however, only 18 % of studies employed quality assurance measures to assess the extent to which the program was delivered as planned. Sixteen and 26 % of studies reported ‘highly’ or ‘moderately’ positive changes (effectiveness) respectively, based on weight loss. Six (16 %) studies reported ‘high’ diabetes risk reduction but ‘low’ to ‘moderate’ weight loss only. Conclusion: Our findings identify that program intensity plays a major role in weight loss outcomes. However, programs that have high uptake—both in terms of good coverage of invitees and their willingness to accept the invitation—can still have considerable impact in lowering diabetes risk in a population, even with a low intensity intervention that only leads to low or moderate weight loss. From a public health perspective, this is an important finding, especially for resource constrained settings. More use of the PIPE framework components will facilitate increased uptake of T2DM prevention programs around the world. Keywords: Implementation, Translational research, Diabetes prevention, Penetration, Implementation, Participation, Effectiveness (PIPE) impact metric, Systematic review, Resource allocation * Correspondence: azizz@student.unimelb.edu.au Melbourne School of Population and Global Health, University of Melbourne, Melbourne 3010, Australia Full list of author information is available at the end of the article © 2015 Aziz et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Aziz et al. Implementation Science (2015) 10:172 Page 2 of 17 Background Another systematic review by Dunkley et al. [27] con- Type 2 diabetes mellitus (T2DM) has emerged as a cluded that pragmatic diabetes prevention programs are major public health challenge with an estimated 387 mil- effective and adherence to international guideline rec- lion people living with T2DM globally [1]. Efforts to ommendations is significantly associated with a greater prevent or delay the onset of diabetes are an urgent weight loss; however, the authors mainly focussed on the public health priority with health, social, and economic effectiveness of the selected translational studies in benefits. Several large randomized controlled trials relation to the main outcome without giving consider- (RCTs) from the US, Finland, China, and India have ation to the adaptability and scalability dimensions of demonstrated that lifestyle interventions can be success- translation. ful in reducing the incidence of T2DM from 29–58 % in A more recent systematic review [28] assessing com- high-risk populations [2–5], with generally good main- bined diet and physical activity promotion programs of tenance for up to 20 years [6, 7]. However, these trials between 3 months and 6 years duration concluded that have mainly focused on efficacy and effectiveness out- more intensive programs resulted in greater weight loss comes. It is now important that more emphasis is given and lowered the risk of T2DM more than less intensive to the implementation and transferability of diabetes programs and this was also found to be cost-effective. prevention programs into real-world settings. ‘Imple- Based on the accumulated evidence from this and other mentation’ research focuses on understanding the pro- reviews, the US Community Preventive Services Task cesses, results, and factors affecting implementation Force (CPSTF) has recently recommended combined under real-world conditions by answering questions diet and physical activity promotion programs to prevent such as ‘why and how interventions work in real-world T2DM [28]. Importantly, however, the above review settings’ [8]. ‘Transferability’ describes the process of ap- examining the translation of diabetes prevention pro- plying the results of research in one situation to other grams like US-DPP did not address issues specifically re- similar situations’ [9]. ‘Real-world’ contexts are settings lated to the long-term scalability and sustainability of where health research findings are applied in practice. such programs. The latter include primary healthcare settings, work or- Another comprehensive review to evaluate the ganizations, churches, and schools. Major transferability generalizability of diabetes prevention trials was con- gaps still remain in translating diabetes prevention from ducted by Whittemore [29] using Glasgow’s[30]reach, ef- research into practice [10]. ficacy, adoption, implementation, maintenance (RE-AIM) Several reviews of diabetes prevention trials have been framework. This is a framework designed to summarize conducted to examine their effectiveness [11–23]. Some how well research trials report on elements related to re- recent reviews have examined how implementation in- search translation. Generalizability is an important elem- fluences effectiveness, most particularly in relation to ent of translational research, i.e. the degree to which the US diabetes prevention program (US-DPP). One of findings from a study or set of studies can be more these reviews [24] summarized lifestyle interventions broadly generalized to populations and settings beyond based on the US-DPP curriculum and described how those in the original studies [31, 32]. This review included different curricula (as a measure of implementation) 16 studies that translated the US-DPP protocol into four affected outcomes. They found that the less-intense distinct settings: (a) hospital outpatient, (b) primary care, version of the US-DPP core curriculum may influence (c) community, and (d) work and church. The author long-term outcomes. Ali et al. [25] conducted a system- found positive outcomes in terms of the reach, efficacy, atic review and meta-analysis of 28 translational studies adoption, implementation, and maintenance of programs. based on the US-DPP and concluded that the utilization However, this review had a narrow focus and only in- of non-medical personnel in delivering diabetes preven- cluded those studies that were based on the US-DPP; no other protocols were considered or included. Another tion interventions can lower the overall costs without compromising effectiveness. Johnson and colleagues limitation was that several studies considered only short- [26] synthesized evidence from translational studies term effectiveness of up to 3 months, and some of the included studies also recruited patients who already had and assessed the impact of interventions delivered outside large randomized trials. The review included diabetes. Laws and colleagues [31] conducted a systematic 17 translational studies from a range of settings and review examining the adequacy of reporting of external concluded that there is potential for less intensive validity components of lifestyle intervention trials interventions to have an impact on future progres- aimed at T2DM prevention. The authors assessed the sion to diabetes in at-risk individuals. However, these generalizability of the findings of 31 studies. Laws re- reviews included studies where subjects had either ported that all studies lacked full reporting on exter- already progressed to diabetes [25, 26] or had short nal validity elements. One of the limitations of this follow-up (e.g. less than 12 months [26]). review was the use of a dichotomous rating scale Aziz et al. Implementation Science (2015) 10:172 Page 3 of 17 (‘reported’ or ‘not reported’) which did not take into Study selection account external validity elements which were re- We included all published studies in the last 15 years ported as continuous measures [31]. (i.e. 2001–2015) that reported on the evaluation of a In summary, a number of recent reviews have focused lifestyle-focused program aimed at individuals at moder- on the real-world effectiveness of T2DM prevention, in ate or high risk of diabetes (e.g. impaired glucose toler- particular, programs derived from the US-DPP. However, ance (IGT), elevated haemoglobin A1c (HbA1c), high these reviews have not systematically examined the body mass index (BMI) or overweight). The inclusion adaptability and scalability dimensions of translation, criteria were adults aged 18 years or older; English lan- that is, what works, under what ‘real-world’ conditions guage publications; and full text. The exclusion criteria and in which contexts; yet this information is critically were studies that were published prior to 2001; that did important for policy makers and program implementers not report at least 1-year follow-up; included partici- who need to identify diabetes prevention programs with pants with known diabetes; included participants not at significant population impact. elevated risk of developing diabetes; and reported on To address this knowledge gap, we have undertaken a multiple intervention components in a single study. systematic review that focuses on identifying the critical Studies were also excluded if they were exclusively ‘diet- success factors for implementing diabetes prevention based’ or ‘exercise-based’ instead of referring to lifestyle programs in real-world settings. We consider the im- or behavioural interventions or if they used lifestyle and portance of program and product design elements that pharmacological interventions. are important to real-world implementation including the effect size, scope of services, scalability potential, Data extraction and long-term sustainability [33]. We use the penetra- An evaluation of T2DM prevention program impact was tion, implementation, participation, effectiveness (PIPE) conducted using the PIPE Impact Metric [33]. The PIPE impact metric, a framework that is highly relevant to im- Impact Metric expresses four elements of program initi- plementation and a formal assessment of the net impact ation and long-term delivery needed to maximize the of health improvement programs that is explicitly linked population impact (i.e., penetration, implementation, to the program design elements noted above [33]. The participation, and effectiveness) as a coefficient. The four key elements of the PIPE framework are as follows: product of all coefficients becomes the PIPE Impact (1) penetration of the program into the population of Metric. For example, a study may report having reached interest, (2) implementation of the proposed set of ser- 250 out of 350 individuals at high risk for T2DM vices, (3) participation in the program, and (4) effective- workers at a company. In this example, penetration is re- ness in generating expected outcomes. Each of the PIPE ported as (250/350) = 0.714 (71 %). Similar coefficients Impact Metric elements may be expressed as a coeffi- are calculated for implementation, participation, and ef- cient, and the product of these four coefficients is re- fectiveness. As numeric data was not available for some ferred to as the PIPE impact metric. The PIPE Impact of the elements in some of the included studies, an alter- Metric can be used to provide feedback to program native coding system, also informed by PIPE, was devel- administrators on gaps in performance, and it also oped for this paper. enables the integration of program design features and identifies where to focus on for performance improve- Data synthesis and analysis ment changes. This paper aims to review the current The coding system is summarized in Table 1. It included evidence about success factors for implementing dia- two steps: (1) an initial scoring and (2) coding of the betes prevention programs in real-world settings using scores into ‘high’, ‘medium’, ‘low’,or ‘not able to calculate the PIPE Impact Metric. (NAC)’ where relevant data was not available. In the ini- tial scoring, coefficients were calculated for penetration and participation. For implementation, the initial scoring Methods was done on three aspects: frequency over the first Data sources and searches 12 months; duration of the entire intervention including A comprehensive search was carried out using PubMed, follow-up contacts; and intervention fidelity. To over- Web of Science, MEDLINE, CENTRAL, and EMBASE come heterogeneity in the kinds of contacts when scor- (February 2014). Search terms were ‘diabetes’ AND ‘pre- ing frequency, a system was developed to standardize vention’ AND (‘program’ OR ‘intervention’) AND (im- the degree of contact based on number, length, and type plementation’ OR ‘translation’). The search was repeated (Table 1). Similarly, for effectiveness, the initial scoring using PubMed to include relevant articles from February was on three criteria: proportion of participants successful 2014 to March 2015. A detailed search strategy is pro- in achieving the main outcome; weight loss (in kilo- vided in Additional file 1. grammes); and diabetes risk reduction (absolute/relative). Aziz et al. Implementation Science (2015) 10:172 Page 4 of 17 Table 1 Development of coding system using PIPE Impact Metric elements PIPE Element Description Coding of the scores Penetration Numerator: the number of individuals reached (invited) ≤33 % = low; 34–66 % = moderate; ≥67 % = high; or NAC (not able to calculate) Denominator: the number of individuals in target population Implementation Implementation was rated on 3 aspects. ≤33 % = low; 34–66 % = moderate; or ≥67 % = high 1. Frequency: the degree of contact (based on number, length, and type) over the first 12 months of an intervention. Different types of contacts were quantified based on the session type in the following way: � 1 group/individual session = 1 session � 1 group/individual session (>3 h) = 2 sessions � 1 online/telephone session = 0.5 session � 1 text/email/fax contact = 0.25 session Numerator: total number of sessions (over the first 12 months) Denominator: 22 (the US-DPP 16 weekly + 6 monthly = 22 sessions) 2. Duration: the duration of the intervention ≤6 months = low; 6–12 months = moderate; and >12 months = high 3. Fidelity: the use of standard curriculum (for example: the No standard curriculum followed = low; a standard curriculum US-DPP) for the delivery of intervention and use of quality was followed but no quality assurance measures were assurance measures to monitor the implementation of the reported = moderate; a standard curriculum was followed intervention and quality assurance measures were applied = high; or NAC (not able to calculate) Participation Numerator: the number of participants enrolled in the intervention ≤33 % = low; 34–66 % = moderate; ≥67 % = high; or NAC (not able to calculate) Denominator: the number of individuals reached (invited) Effectiveness Effectiveness was rated on 3 criteria: ≤25 % = low; 26–40 % = moderate; >40 % = high; or NAC (where information is not provided) 1. Success criterion/proportion of successful participants: Numerator: participants who achieved the main outcome (i.e. weight loss ≥5%) Denominator: total number of participants enrolled in the intervention/total number of participants completed 12-month measurements 2. Average weight loss: the average weight loss (in kilogrammes) ≤2.3 kg = low; 2.4–4.6 kg = moderate; >4.6 kg = high; or NAC (where information is not provided) 3. Risk reduction: diabetes risk reduction (absolute/relative) Risk reduction: ≤15 % = low; 16–30 % = moderate; >30 % = high; or NAC (where information is not provided) The coding of the scores (into high, medium, low, and assessed for eligibility. A total of 76 articles from 38 NAC) for each element was agreed-upon by the review studies were included in the review. Table 2 describes team, and ratings were independently derived by three re- the characteristics of the included studies. searchers (ZA, PA, JO). The ratings of all studies were Studies that met the eligibility criteria for this review then reviewed by all authors, and when disagreement on were mainly based on either the US-DPP or Finnish-DPS. rating occurred, they were resolved with consensus among Seventeen (45 %) studies were implemented in the USA, 4 all authors. in the UK, 4 in the Netherlands, 8 in other European countries, 3 in Australia, and 2 in Japan. There were no Results studies from low and middle income countries that met A detailed PRISMA flow diagram is attached (Fig. 1). our eligibility criteria. The sample size for the participants The initial literature search (February 2014) returned enrolled in each of these studies ranged between 8 and 2992 publications and 61 additional articles were identi- 2553 participants. The studies were conducted in a range fied through hand searching of references from the bibli- of settings including primary health care, faith-based, ographies of articles identified. Two thousand thirty- workplace, and other community-based settings. nine articles were screened after removing duplicates. Each study was assessed on the four components of An additional 5 articles were included after updating the PIPE Impact Metric using a coding system described search until March 2015. A total of 180 articles were in Table 1. Table 3 describes the ratings of all Aziz et al. Implementation Science (2015) 10:172 Page 5 of 17 Records identified Additional articles through database identified through searching hand searching (n = 2,992) (n=61) Records after duplicates removed (n = 2,039) Records excluded by Records screened title/abstract (n = 2,039) (n = 1,864) Additional studies (from February 2014 to March 2015) (n = 5) Full-text articles excluded, with reasons Full-text articles n = 104 (92 studies) assessed for eligibility Reasons: (n = 180) * < 12-months follow-up (n = 19) * Included people with T2DM (n = 19) * Study protocol/design (n = 13) * Included young/healthy people (n = 9) * Included multiple interventions (n = 7) * No full article found (n = 7) Full-text articles * Published prior to 2001 (n = 3) included in this review * Other (n = 15) n = 76 articles (38 studies) Fig. 1 PRISMA flow diagram included studies based on the elements of the PIPE only study rated as having ‘moderate’ penetration, was a Impact Metric. church-based study [39]. The church roster included 407 members, whereas 37 % of adults (approximately Program penetration 150) who attended Sunday gatherings were invited to Our analysis shows that only nine (24 %) studies re- complete a diabetes risk assessment. The study rated as ported their estimated target population, from which having ‘low’ penetration in our analysis [40] included 18 only seven studies reported the proportion of the target participating centres that covered all primary care ser- population that was reached with invitations to engage vices for 4.5 % of the population in Catalonia, out of in the program or intervention. Out of these seven stud- which less than 1 % of the population was invited for ies, five had ‘high’, one had ‘moderate’, and one had ‘low’ screening. penetration into their target populations (see Table 3 and Additional file 2: Table S4). Target populations in- Program implementation cluded patients, employees, and church attendees. Each In order to assess implementation, we evaluated the de- study used various strategies to recruit potential partici- gree of contact (based on number, length, and type) dur- pants including mail invitations, posted flyers, advertis- ing the first year of the intervention as frequency; the ing through media, contacting local physicians, local duration of the entire intervention; and the fidelity of churches, or using intranet or work meetings in the the intervention (see Table 1). All studies in our analysis workplace setting. reported on frequency and duration. Thirty-four percent The five studies that were rated as having ‘high’ pene- of all studies implemented ‘high’ frequency interven- tration in our analysis applied heterogeneous strategies tions, and 39 % studies delivered intervention over the to reach their target group. Two studies contacted a pre- period of 12 months or more. The number of contacts defined group of people at risk: one at worksite, where varied from a single small group session to 32 group ses- all employees who had above average waist circumfer- sions. About two thirds (66 %) of the programs based on ence were invited for screening [34], and in the other the US-DPP model adopted a ‘low-’ to ‘moderately in- study [35] all eligible subjects with high risk for glucose tense’ version (based on the degree of contact) as com- intolerance from a cohort representing general popula- pared to the original. The main adaptation was the tion were contacted. Three studies contacted 100 % [36], reduction from 16 to fewer sessions. In addition, groups 68 % [37], and 70 % [38] of the target population for se- led by volunteers as opposed to health care professionals lective screening either by mail or by appointment. The and use of telephone as opposed to face-to-face delivery Included Eligibility Screening Identification Aziz et al. Implementation Science (2015) 10:172 Page 6 of 17 Table 2 Characteristics of the included studies Year Author Study ID Country Setting Study population Sample Study design Intervention size 2003 Mensink et al. SLIM Netherlands Unclear Adults at risk of T2DM 114 RCT 3 individual and 1 group session during 1 year + [35] participants were encouraged to participate in the exercise program 3 times a year 2005 Kosaka et al. Japanese DPP Japan Hospital-based Adults with IGT 458 RCT Detailed instructions on lifestyle were repeated every [62] 3 to 4 months during hospital visits 2006 Oldroyd et al. Newcastle lifestyle UK Primary care Adults with IGT 78 RCT 12 individual 15- to 20-min review appointments over [63] intervention 24 months (6 in the first 6 months, 1 after 9 months and 5 at 2 monthly intervals between 12 and 24 months) 2007 Absetz et al. GOAL LIT Finland Primary care Adults at risk of T2DM 352 Before and after Six 2-hourly group counselling sessions over 8 months [41] 2007 Bo et al. [36] Italian Trial Italy Primary care Adults with metabolic syndrome 335 RCT 1 individual and four 1-hourly group sessions 2007 Davis-Smith DPP (church- USA Community Adults at risk of T2DM 10 Before and after 6-session lifestyle intervention delivered over a 7 week et al. [39] based) (church) period 2007 Laatikainen Greater Green Australia Primary care Adults at risk of T2DM 311 Before and after 6 structured 90-min group sessions delivered during et al. [64] Triangle (GGT) an 8-month period 2008 Ackermann DEPLOY USA Community Adults at risk of T2DM 92 RCT Sixteen 1 to 1.5-hourly small group sessions over 16 to et al. [65] (YMCA) 20 weeks and monthly large-group meetings 2008 Boltri et al. DPP (church- USA Community Adults with pre-T2DM 8 Before and after 16 weekly group sessions conducted over 4 months [66] based) (church) 2008 Payne et al. BDPPI Australia Outpatient setting Adults at risk of T2DM 122 Before and after 6-week group self-management education program, [67] 12-week gym- or home-based resistance training, and three 2-h group reinforcement sessions during 34-week maintenance program 2009 Kramer et al. GLB (2007–2009) USA Community Adults with pre-T2DM 42 Before and after 12 weekly sessions (~60 min) and participants were [68] offered the opportunity to attend monthly support meetings for 9 months after completion of the intervention 2009 Kulzer et al. PREDIAS Germany Outpatient setting Adults at risk of T2DM 182 RCT 12 lessons lasting ~90 min each [69] 2009 Penn et al. EDIPS—Newcastle UK Outpatient setting Adults with IGT 102 RCT A 30-min session immediately following randomisation [70] and 2 weeks later, then monthly for the first 3 months and every 3 months thereafter up to 5 years 2010 Almeida et al. Colorado weight USA Integrated health Adults with pre- T2DM 1520 Matched cohort A single 90-min small group session [71] loss intervention care organization 2010 Makrilakis et al. DE-PLAN Greece Greece Primary care Adults at risk of T2DM 191 Before and after 6 sessions (1 h each) held by a registered dietician in [72] (workplace) the groups of 6 to 10 persons 2010 Parikh et al. Project HEED USA Community Adults at risk of T2DM 99 RCT A peer-led lifestyle intervention group, presented in a [73] workshop consisting of eight 1.5-h sessions over 10 weeks Aziz et al. Implementation Science (2015) 10:172 Page 7 of 17 Table 2 Characteristics of the included studies (Continued) 2010 Vanderwood Montana CVD and USA Health care Adults at risk of T2DM and CVD 355 Before and after 16 weekly group sessions and 6 monthly group et al. [45] DPP facilities (pilot study) sessions 2010 Vermunt et al. APHRODITE Netherlands Primary care Adults at risk of T2DM 925 RCT 11 consultations of 20-min over 2.5 years, five 1-h [74] group meetings and 1-h personal consultation with the dietician 2011 Boltri et al. DPP (church- USA Community Adults with pre-T2DM 37 Before and after 6 or 16 weekly group sessions [75] based) (church) 2011 Gilis- DE-PLAN Poland Poland Primary care Adults at risk of T2DM 175 Prospective cohort 10 group sessions lasting for 4 months, 6 telephone Januszewska motivation sessions, and 2 motivation letters sent to et al. [76] participant 2011 Katula et al. HELP PD USA Community Adults with pre-T2DM 301 RCT ~26 weekly group sessions for the first 6 months, 3 [77] (various venues) personalized consultations with a registered dietician, 18 monthly group sessions, and monthly phone contact for the last 18 months 2011 Kumanyika Think health! USA Primary care Adults with high BMI and 261 RCT Brief monthly contact with a lifestyle coach (LC) for et al. [48] weight 12 months and 10–15 min counselling sessions with primary care providers every 4 months. Bi-monthly sessions with LC for the second year 2011 Nilsen et al. Nilsen et al. Norway Primary care Adults at risk of T2DM 213 RCT The individual and interdisciplinary group participated [42] in a group-based program, 1 day (5 h per day) each week for 6 weeks An individual 30-min consultation with a nurse or ergonomist completed the intervention 1 month after the last group meeting 2011 Penn et al. NLNY UK Leisure and Adults at risk of T2DM 218 Before and after A 10-week program of twice-weekly 1.5-h sessions, [43] community followed by ongoing support with regular mobile settings phone text message and email reminders, ‘drop-in’ activity sessions continued up to 12 months 2011 Ruggiero et al. HLP USA Community Adults at risk of T2DM 69 Before and after 16 weekly core sessions and 6 monthly after-core [78] (various venues) sessions 2011 Sakane et al. Japanese Study Japan Primary care Adults with IGT 304 RCT 4 group sessions of 2 to 3 h (for the first 6 months), [79] (workplace) individual sessions twice a year for 3 years. Between-visit contact by fax was also made monthly during the initial 12 months 2012 Costa et al. DE-PLAN-CAT Spain Primary care Adults at risk of T2DM 552 Prospective cohort A 6-h educational program (scheduled in 2 to 4 [40] individual/small group sessions), and regular contact by phone or text message for at least once every 6 to 8 weeks 2012 Janus et al. pMDPS Australia Primary care Adults at risk of T2DM 92 RCT 6 structured 90-min group sessions. The first 5 sessions [46] were at 2 weeks intervals and the final session was 8 months after the first 2012 Kanaya et al. LWBW USA Community Adults at risk of T2DM 238 RCT The intervention was primarily telephone-based [50] counselling (12 calls) with 2 in-person sessions Aziz et al. Implementation Science (2015) 10:172 Page 8 of 17 Table 2 Characteristics of the included studies (Continued) and 5 optional group workshops over 1 year period 2012 Lakerveld et al. Hoorn Prevention Netherlands Primary care Adults at risk of T2DM and/or 622 RCT Six individual 30-min counselling sessions, [37] Study CVD followed by 3-monthly booster sessions by phone for a period of 1 year. 2012 Ockene et al. LLDPP USA Community Adults at risk of T2DM 312 RCT 3 individual and 13 group sessions over a [80] 12 month period 2012 Piatt et al. [81] GLB (2005–2008) USA Community Adults with metabolic syndrome 105 Before and after 12 weekly sessions over 12 to 14 weeks (lasted ~90 min) in the groups of 5 to 13 participants 2013 Jiang et al. SDPI-DP USA Community Adults with pre-T2DM 2553 Before and after 16 group sessions in the first 16 to 24 [82] weeks and monthly individual lifestyle coaching sessions 2013 Ma J et al. [38] E-LITE USA Primary care Overweight/obese adults with 241 RCT 12 weekly group sessions (1.5 to 2 h each) in increased cardiometabolic risk the first 3 months. From month 4 to 15, contact every 2 to 4 weeks depending on participant needs and preferences. Individual, secure email/ phone contacts with personalized progress feedback and lifestyle coaching throughout the maintenance phase (month 4 to 15) 2014 Duijzer et al. SLIMMER Netherlands Primary care Adults at risk of T2DM 31 One group pre-test In addition to 6 individual consultations [49] post-test (in total 4 h per participant), on average, participants received 5.2 consultations by dieticians and 34.1 sports lessons 2014 Sepah et al. Prevent USA Online platform Adults with pre-T2DM 220 Quasi-experimental 16 online weekly lessons. Participants were then [47] research design offered to continue with a post-core lifestyle change maintenance intervention, with the entire intervention (core plus post-core) totalling 12 months 2014 Zyriax et al. DELIGHT Germany Primary care Adults at risk of T2DM 241 Before and after 12 weekly sessions (for the first 6 months), 6 monthly [34] (workplace) and 6 biweekly sessions (for the next 6 months). For year 2 and 3 quarterly 1.5-h sessions 2015 Savas et al. IGT care call UK Primary care Individuals with IGT 55 Observational A telephone service providing a 6 month lifestyle [44] study education program (20 min × 6), in addition to an introduction call (10 min) and action planning call (40 min) Aziz et al. Implementation Science (2015) 10:172 Page 9 of 17 Table 3 Scoring for each PIPE element by study Author Year Study Penetration Implementation Participation Effectiveness Frequency Duration Fidelity Success rate Weight loss Risk reduction (absolute/relative) Mensink et al. 2003 [35] SLIM High Low High Low Low NAC Moderate High Kosaka et al. 2005 [62] Japanese DPP NAC Low High Moderate NAC NAC Moderate High Oldroyd et al. 2006 [63] Newcastle LI NAC Moderate High NAC Low NAC Low NR Absetz et al. 2007 [41] GOAL LIT NAC Moderate Moderate Moderate High NAC Low NR Bo et al. 2007 [36] Italian Trial High Low High High Low NAC Low High Davis-Smith et al. 2007 [39] DPP (church-based) Moderate Low Low Moderate Low NAC High NR Laatikainen et al. 2007 [64] GGT NAC Low Moderate High Low NAC Moderate NR Ackermann et al. 2008 [65] DEPLOY NAC High High Moderate Low NAC High NR Boltri et al. 2008 [66] DPP (church-based) NAC High Low Moderate Low NAC Low NR Payne et al. 2008 [67] BDPPI NAC High High Moderate NAC Moderate Moderate NR Kramer et al. 2009 [68] GLB (2007 – 2009) NAC High Moderate Moderate Low NAC High NR Kulzer et al. 2009 [69] PREDIAS NAC Moderate Low Moderate NAC NAC Moderate NR Penn et al. 2009 [70] EDIPS- Newcastle NAC Moderate High NAC Low NAC Low High Almeida et al. 2010 [71] Colorado weight loss intervention NAC Low Low NAC Low Low Low NR Makrilakis et al. 2010 [72] DE-PLAN Greece NAC Low Moderate Low Low NAC Low NR Parikh et al. 2010 [73] Project HEED NAC Moderate Low Moderate Low Moderate Moderate NR Vanderwood et al. 2010 [45] Montana CDDP NAC High Moderate Moderate High High High NR Vermunt et al. 2010 [74] APHRODITE NAC Moderate High NAC Low NAC NAC NR Boltri et al. 2011 [75] DPP (church-based) NAC Low (2 churches) Low Moderate Low NAC Low NR High (3 churches) Gilis-Januszewska et al. 2011 [76] DE-PLAN Poland NAC Moderate Moderate Low Low Low Low NR Katula et al. 2011 [77] HELP PD NAC High High High Low NAC High NR Kumanyika et al. 2011 [48] Think Health! NAC Moderate Moderate Moderate Moderate Low Low NR Nilsen et al. 2011 [42] Nilsen et al. NAC High High NAC High Moderate NAC NR Penn et al. 2011 [43] NLNY NAC High Moderate Low High Low Low NR Ruggiero et al. 2011 [78] HLP NAC High Moderate Moderate Low Moderate Low NR Sakane et al. 2011 [79] Japanese Study NAC Moderate High Moderate Low NAC Low High Costa et al. 2012 [40] DE-PLAN-CAT Low Low High NAC Low NAC Low High Janus et al. 2012 [46] pMDPS NAC Low Moderate High High NAC Moderate Moderate Kanaya et al. 2012 [50] LWBW NAC Moderate Moderate Moderate Moderate NAC Low NR Lakerveld et al. 2012 [37] Hoorn Prevention Study High Moderate Moderate High Low NAC NAC NR Aziz et al. Implementation Science (2015) 10:172 Page 10 of 17 Table 3 Scoring for each PIPE element by study (Continued) Ockene et al. 2012 [80] LLDPP NAC High Moderate Moderate Low NAC Low NR Piatt et al. 2012 [81] GLB (2005–2008) NAC Moderate Moderate Moderate Low Moderate NAC NR Jiang et al. 2013 [82] SDPI-DP NAC High Moderate Moderate Low NAC Moderate NR Ma J et al. 2013 [38] E-LITE High High High High Low High High NR Duijzer et al. 2014 [49] SLIMMER NAC Low Moderate Moderate Moderate NAC Moderate NR Sepah et al. 2014 [47] Prevent NAC Moderate High Moderate High High NAC NR Zyriax et al. 2014 [34] DELIGHT High High High Low Low NAC NAC NR Savas et al. 2015 [44] IGT Care Call NAC Low Moderate High High NAC Moderate NR Details on the scoring of all included studies based on the elements of the PIPE Impact Metric framework are provided in Additional file 2: Table S4–Table S7 NAC not able to calculate, NR not reported Aziz et al. Implementation Science (2015) 10:172 Page 11 of 17 of individual sessions were also frequently observed ad- Also, none of the seven studies that were rated as ‘high’ aptations. Only a small proportion (16 %) of studies re- participation provided enough information on reaching ported ‘low’ duration, i.e. intervention delivered over the out to their target populations, and hence, penetration period of 6 months or less. could not be calculated. However, the information avail- Implementation fidelity was defined as intervention able suggests that the studies where high-risk partici- being based on a standard curriculum for example the pants were identified and referred through their GPs or US-DPP and whether any quality assurance measures nurses resulted in ‘high’ participation rates. were applied to monitor the implementation of the intervention. Twenty-seven (71 %) studies were based Program effectiveness on a standard curriculum, out of which only 7 studies Effectiveness was rated based on three criteria: propor- had quality assurance measures applied to monitor the tion of successful participants; average weight loss; and implementation. In some of these studies, while authors diabetes risk reduction (absolute/relative) (Refer to reported on the efforts made to minimize the potential Table 1). None of the studies reported on all three cri- lack of fidelity, none of these studies provided informa- teria. Seventeen (45 %) studies reported the use of tion on the extent to which the various components intent-to-treat analysis; however, for the purpose of our were delivered except one where the program compo- analysis, effectiveness indicators were considered as pre- nents were more frequently added (40 %) than omitted sented in each of the studies. (28 %) [41]. Refer to Table 3 and Additional file 2: Table One third (n = 12, 32 %) of the studies reported the S5 for further details. proportion of successful participants who achieved the primary outcome (i.e. 5 % weight loss). The proportion Program participation of successful participants ranged between 20 and 64 %. The majority of the studies in our analysis (n = 35; 92 %) Thirty-two (84 %) studies reported average weight loss reported participation. Twenty-five (71 %) of these 35 by participants at 12 months, with a range from 0.45 to studies achieved ‘low’ participation rates. Half of these 7.7 kg. Only six of these studies were rated ‘high’ where studies achieved participation rates equivalent to or average weight loss by participants was more than lower than 10 %. Only 7 (18 %) studies had ‘high’ partici- 4.6 kg. Sixteen (42 %) studies were rated ‘low’ on the pation rates. For four of these studies [41–44], partici- basis of average weight loss of ≤2.3 kg. Only seven pants were recruited by referral from physicians, general (18 %) studies reported the data on diabetes risk re- practitioners (GPs), or nurses from the participating duction (absolute/relative), where six studies were health facilities and invited to attend a screening clinic. rated ‘high’ and one was rated ‘moderate’.Scores Two of the 7 studies [45, 46] used a combination of could not be calculated in three studies due to the strategies to recruit potential participants including con- lack of numerical data. tacting local physicians and primary healthcare practices; Thirteen studies (34 %) were rated ‘high’ on at least advertising through media; and recruiting through local one of the three criteria. One study [45] that was rated employers, work sites, churches, and service groups. ‘high’ on both success rate and weight loss, reported that One study [47] recruited participants from online ad- 64 % of participants achieved 5 % weight loss goal with vertisement, seeking individuals with a self-reported the average 12-month weight loss of 7.7 kg. This study clinical diagnosis of prediabetes occurring within the was not only ‘highly’ effective but also had ‘high’ partici- past year; however, recruitment was based on self- pation rates. The intervention included ‘high’ number of selection by participants, which does not reflect a sessions with ‘moderate’ duration and ‘moderate’ fidelity. truly random sample [47]. Another study [38] that had ‘high’ penetration into its Three studies were scored as having ‘moderate’ partici- target population appeared to be ‘highly’ effective based pation, where 63 % [48], 57 % [49], and 44 % [50] of the on the proportion of successful participants (50 %) and individuals were enrolled after assessing all ‘invited’ indi- average weight loss (6.3 kg at 15 months). This study viduals, for study eligibility (see Table 3 and Additional however, reported ‘low’ participation and delivered file 2: Table S6). In most of the remaining studies, ‘high’ number of sessions and had ‘high’ duration of the ‘low’ participation was attributed to either the intervention. non-eligibility of potential participants or the refusal Five out of six studies that reported ‘high’ weight loss to participate. delivered ‘high’ number of sessions. For studies where information was available for both Out of 10 studies that reported ‘moderate’ weight loss, penetration and participation, it was observed that ‘high’ 6 (60 %) implemented interventions with ‘low’ number penetration into the target population did not have posi- of sessions/contacts. Sixteen (42 %) studies had ‘low’ ef- tive effect on participation. All five studies rated as ‘high’ fectiveness based on average weight loss, out of these, 8 penetration in the analysis, reported ‘low’ participation. (50 %) studies delivered ‘moderate’ number of sessions, Aziz et al. Implementation Science (2015) 10:172 Page 12 of 17 whereas 4 (25 %) delivered ‘high’ and 4 (25 %) delivered ‘high’ penetration resulted in either ‘high’ weight loss ‘low’ number of sessions. Six studies reported ‘high’ ef- or ‘high’ diabetes risk reduction. Hence, scalability of fectiveness in risk reduction despite ‘low’ (4 studies) or the program to reach a large audience appears to be ‘moderate’ (2 studies) effectiveness in weight loss, and an important ingredient for population-level impact. these were also all studies with only ‘low’ (4 studies) or In summary, while an intensive intervention plays an ‘moderate’ (2 studies) frequency, but with ‘high’ duration important role in achieving successful weight loss out- (6 studies). See Table 3 and Additional file 2: Table S7 comes, highly scalable moderate- to low-frequency inter- for more details. ventions appear to have major potential to achieve diabetes risk reduction in populations. Discussion From a program implementation perspective, it is im- This is the first systematic review to evaluate the imple- portant to clearly define and estimate the size of the tar- mentation of real-world diabetes prevention programs get population. Without a reasonable estimate of the using the PIPE Impact Metric framework that deploys size of the target population, there is a risk that the pro- four highly relevant elements for monitoring program gram will not be scalable or sustainable [33]. We found impact in real-world settings. As such, this review com- estimates of target populations to be reported by less plements other recent reviews, e.g. Dunkley et al. 2014 than one third of included studies. Hence, it is possible [27], by providing a more detailed understanding of key that many of these programs were delivered to highly se- factors underlying successful translation and implemen- lected populations which limit their generalizability. tation of diabetes prevention programs in real-world We used the US-DPP intervention as a benchmark to contexts. We have also defined the specific scope of assess the intensity of dose-delivered (including fre- services for calculating the overall costs of services being quency, duration, and fidelity) in our included studies. provided. From both an organizational and societal In order to deal with heterogeneity in the types of con- perspectives, these issues are important to consider since tacts, we constructed a framework to standardize the the relative costs and benefits of such services and degree of contact based on number, length, and type. programs are important determinants of their uptake We found that many programs adopted a ‘low-’ to ‘mod- and adoption. Our review of studies published over erately intense’ version, compared to the US-DPP inter- the last 15 years aims to identify the components of vention. In practice, this also means that many of the diabetes prevention programs with the highest popu- components were either omitted or modified from the lation impact. original US-DPP curriculum. As stated earlier, we also Our review highlights several important findings. First, note that ‘high’-frequency interventions with ‘high’ to confirming earlier reviews, our analysis demonstrates ‘moderate’ duration and fidelity were associated with that lifestyle-focused diabetes prevention programs that greater weight loss. This is consistent with a systematic have a ‘high’ degree of contact have more potential to review and meta-analysis by Dunkley et al. [27], where achieve effective outcomes, especially when measured by reviewers coded intervention content based on the weight loss. These programs have typically been based recommendations for lifestyle interventions for the pre- on the US-DPP model and have used a very structured vention of diabetes provided by both the European protocol to maximize program fidelity. However, the Guideline and Training Standard for Diabetes Preven- problem with this approach is that in these studies, tion (IMAGE) project [51] and National Institute for program participation tends to be quite low; and fur- Health and Care Excellence (NICE) [52] and found that thermore, none of these studies reported diabetes risk adherence to guidelines on the content and delivery was reduction. significantly associated with a greater weight loss. Our Second, six of the studies showed different degrees of findings on the potential of ‘low’ to ‘moderate’ frequency effectiveness for different outcomes. For example, dia- interventions with longer duration to achieve significant betes risk reduction could be ‘high’ even when effective- risk reduction support earlier findings [25, 26]. A recent ness in weight loss was ‘low’ or ‘moderate’. Surprisingly, CPSTF review [28] shows lower weight loss than the these were all studies of ‘low’ or ‘moderate’ frequency, US-DPP but still concluded strong evidence of effective- but ‘high’ duration. This could be very promising espe- ness. The recommendations from this review were fur- cially for settings where intervention resources are con- ther supported by an effectiveness and economic review strained but when large populations can be reached by [53, 54]. So, in general, it seems that studies beyond the such programs. original US-DPP generate somewhat lower effect for Third, we found that ‘high’ penetration into the tar- weight loss but still generate meaningful positive impact get population with invitations to engage prospective on the reduction of incidence of T2DM. Lindstrom et al. participants in the program do not necessarily result have also previously discussed the positive effects of de- in ‘high’ participation. However, three studies with creased fat and increased fibre intake on diabetes risk Aziz et al. Implementation Science (2015) 10:172 Page 13 of 17 reduction in the absence of weight loss [55]. We also used either criterion—diabetes risk reduction or weight know that in many populations diabetes risk can be loss—to assess effectiveness. high at lower levels of weight, with other factors be- We found that weight loss was reported in several dif- sides weight loss playing a critical role in risk reduc- ferent ways. For example, the percentage weight lost tion [56, 57]. could be interpreted to mean weight loss to percentage Examining the implementation component further, in achieving a particular weight loss target or the average calculating ‘frequency’, we have used contacts made in percentage of weight that subject lost. One study [45] re- the initial 12 months only because most of the studies ported both 5 and 7 % weight loss, whereas another [40] did not extend beyond 12 months. In those that did, the reported the proportion of successful participants who initial 12 months can be considered as the ‘action’ phase, achieved at least 3 % weight loss. For consistency, in our bringing about the lifestyle changes, and beyond that is a review, we only used 5 % weight loss when it was re- follow-up and maintenance phase, which some studies ported in several different ways. support with less frequent contacts. However, examination of studies with ‘low’ effective- In translational research, a systematic evaluation of ness reveals that for some of the studies the reported program fidelity is important to assess the extent to changes in weight loss were very small. Some of these which program was implemented as designed. We based studies reported a significant reduction in weight follow- our definition of fidelity on whether a standard curricu- ing the active intervention phase, but the weight was lum was used to guide the delivery of intervention and partly or, in some studies, entirely regained by the end of whether quality assurance measures were placed to 12 months. Lack of consistency in the way weight loss monitor the implementation of the intervention. Not outcomes are reported and analysed needs to be ad- many studies clearly reported whether the quality as- dressed in future translational research [29]. Cardona- surancemeasureswereabletoguarantee ‘high’ fidel- Morrell and colleagues [21] suggests the establishment ity, this clearly being one of the next important steps of a registry of translational projects using consistent, in program development. This needs to be examined measurable outcomes to add more certainty to effective- in future studies. ness analyses. All but three studies in our review reported participa- Our review includes diabetes prevention translational tion, and only seven studies scored ‘high’ on participa- programs published since 2001 and until 2015. The tion rates. The high participation may be a reflection of studies included in this review have implemented ‘high’ highly targeted penetration, but because of unavailability (34 %), ‘moderate’ (37 %), and ‘low’ (29 %) frequency in- of information, we could not calculate penetration coef- terventions; however, we noticed that most of the ‘low- ficients for any of these seven studies. However, the in- frequency’ interventions were conducted in earlier years, formation available suggests that the studies where high- whereas, designing ‘moderate’ to ‘high’ intensity inter- risk participants were identified and referred through ventions occurred in more recent years. However, ‘par- their GPs or nurses resulted in ‘high’ participation rates, ticipation’ has been consistently low in a majority of the underlying the important role of providers. The high studies over the last 15 years. One of the reasons for this participation could also be a reflection of high motiv- may be the fact that program planners focus on the con- ation for change among at-risk individuals, a factor that tent of the interventions instead of balancing the con- the providers can further enhance in face-to-face con- tent with the experience of the participant—that is, on tact, e.g. by applying strategies from motivational inter- the ‘participation’ dimension and the engagement factor viewing [58]. In our review, of the seven studies that that connects the participant with the intervention. scored ‘high’ for participation, five studies reported ‘high’ Future translational research in this field needs to in- to ‘moderate’ effectiveness based on success rate and/or vest in designing recruitment more carefully to ensure weight loss. high program reach; examining factors that optimize The PIPE Impact Metric elements are interrelated in engagement and retention in the structured lifestyle pro- that participation is always a proportion of penetration grams; and maximizing adherence to the long-term be- and effectiveness can only be attributed to those who haviour changes [59]. participated. Effectiveness, in this context, is defined as the number or proportion of participating cases who Limitations reached a priori defined success criterion. In prevention Several studies did not provide relevant numeric data to of T2DM, diabetes risk reduction is one such success allow the calculation of PIPE coefficients. Hence we criterion. In many studies, however, weight loss was also were unable to compare the overall program impact of a main outcome—either in individual cases as a percent- the included studies. Additionally, any successfully trans- age of overall body weight or across a population as an lated diabetes program should ideally be accessible to those average percentage of weight loss. In our review, we most in need and should have some clear relationship with Aziz et al. Implementation Science (2015) 10:172 Page 14 of 17 the health care delivery system, and we were not able to and program implementation strategies for targeting evaluate these two elements in most of the studies included both reach and program participation that also empha- in this review. sizes long-term program adherence. In our analysis of implementation, we only considered To improve the translation of diabetes prevention pro- standard sessions and have not included the extracurric- grams in real-world settings, we suggest a more rigorous ular or optional activities offered by some of the pro- reporting of program elements and components to grams that may have impact on the future adoption of evaluate these programs to assess the practical value these kinds of programs. This was mainly due to the het- [27] of the diabetes prevention programs. In particular, erogeneity in how the extracurricular activities were or- more detailed reporting on the four key PIPE Metric ganized and also how they were reported. However, we components will provide important insights and has the do acknowledge that these ‘spinoff’ activities are likely to potential to facilitate increased uptake of T2DM prevention be very important for sustainability and the wider diffu- programs worldwide. We also suggest a greater consistency sion of programs, their maintenance, and sustainability. of reporting main outcomes and a standardization of Although there has been a lot of emphasis on the need reporting criteria for translational diabetes prevention pro- for translating T2DM programs in low and middle in- grams implemented in real-world settings. come countries where the majority of people in the world at high risk of diabetes and its progression live Conclusions [60], we did not find any studies from such countries Our findings based on program implementation over the that met our eligibility criteria. The Kerala Diabetes Pre- period of 15 years suggest that while a high-frequency vention Program [61] in India is one of the first such im- intervention plays an important role in achieving high plementation trials to evaluate a peer-led, group-based weight loss outcomes, programs with ‘low’-intense inter- lifestyle intervention program (based on the Finnish ventions have also shown high reductions in the inci- GOAL study) [41] among individuals at high risk of de- dence of T2DM. This suggests that even when the veloping T2DM in rural India. The trial is currently be- effectiveness of an intervention is moderate in terms of ing implemented and will provide important reference weight loss, it can have a profound impact on the devel- to the translation of diabetes prevention programs in opment of a disease at the population level—provided India and similar countries. enough effort is put into guaranteeing high penetration Also, we did not find any studies that have previously and participation as well. From a translation perspective, utilized the PIPE Impact Metric framework in diabetes not many studies provide the necessary information to prevention. More research is needed to understand and estimate the overall impact of such programs. Key ele- apply the few but essential elements of the PIPE model ments of the PIPE Impact Metric are not routinely re- to measure the overall public health impact of diabetes ported in many published implementation trials of prevention interventions. diabetes prevention which therefore reduces their utility for information resource allocation and ‘real-world’ im- Implications for practice plementation. More rigorous evaluation methods are re- The original US-DPP [2] and Finnish-DPS [3] efficacy quired to better understand the factors that influence trials demonstrated that lifestyle intervention is an ef- the likely success of such interventions in the future. fective way to reduce the risk of T2DM in high-risk adults. However, achieving better translation of these Additional files programs still remains challenging after 15 years of re- search. Our findings suggest that program planners and Additional file 1: Search Strategy. Search strategy used for identifying implementers should aim to design high-intensity pro- included studies. (PDF 371 KB) gram with frequent contacts if the primary target is Additional file 2: Table S4–Table S7 Detailed scoring of all weight loss. However, if the primary aim is diabetes risk included studies based on the elements of the PIPE Impact Metric reduction, this can also be achieved with lower fre- framework. (PDF 772 KB) quency of contacts but with a program duration of at least 12 months. With this program design, program Competing interests planners should expect only low or moderate weight The authors declare that they have no competing interests. loss. To have a broader public health impact, programs with lower frequency of contacts but with a program Authors’ contributions duration of at least 12 months might be more feasible The ratings of the included studies were independently derived by three authors (ZA, PA, JO). The ratings of all studies were then reviewed by all but this requires program strategies that simultaneously authors and when disagreement on rating occurred they were resolved with address both penetration and participation. Future trans- consensus among all authors. ZA drafted the manuscript. All authors read, lational research needs to identify effective recruitment provided feedback, and approved the final manuscript. Aziz et al. Implementation Science (2015) 10:172 Page 15 of 17 Authors’ information subjects with impaired glucose tolerance (IDPP-1). Diabetologia. Zahra Aziz MPH is a PhD candidate in the School of Population and Global 2006;49(2):289–97. Health University of Melbourne in Australia. Her PhD is on an evaluation of 6. Diabetes Prevention Program Research Group. 10-year follow-up of diabetes the translation of diabetes prevention programs in ‘real-world’ settings with incidence and weight loss in the diabetes prevention program outcomes a particular focus on the National Health and Medical Research Council study. The Lancet. 2009;374(9702):1677–86. (NHMRC) funded study titled the Kerala Diabetes Prevention program based 7. Li G, Zhang P, Wang J, Gregg EW, Yang W, Gong Q, et al. The long-term effect in rural India. She is also involved in evaluating the Australasian Peers for of lifestyle interventions to prevent diabetes in the China Da Qing diabetes Progress trial in Victoria, Australia. Before commencing her PhD, she has prevention study: a 20-year follow-up study. Lancet. 2008;371(9626):1783–9. worked as an evaluation specialist with the Aga Khan Health Service, 8. Peters DH, Adam T, Alonge O, Agyepong IA, Tran N. Implementation Pakistan for 5 years and has also completed a Master of Public Health research: what it is and how to do it. Bmj. 2013;347:f6753. degree from the University of Melbourne, Australia. She has led the 9. Jensen D. In: Given LM, editor. “Transferability.” The SAGE encyclopedia of conducting of this systematic review. qualitative research methods. Thousand Oaks: SAGE Publications, Inc; Pilvikki Absetz PhD is an Adjunct Professor of Health Promotion at University 2008. p. 887. SAGE knowledge. Web. 12 Mar. 2015. of Tampere, School of Health Sciences, Finland. She is a global expert in lifestyle 10. Yates T, Davies M, Khunti K. Preventing type 2 diabetes: can we make the interventions in prevention and management of type 2 diabetes and other evidence work? Postgraduate medical journal. 2009;85(1007):475–80. chronic conditions. She has been a lead or co-investigator of the following 11. Yamaoka K, Tango T. Efficacy of lifestyle education to prevent type 2 major studies that are directly relevant to type 2 diabetes: the Finnish GOAL diabetes: a meta-analysis of randomized controlled trials. Diabetes care. Lifestyle Implementation Trial; the TERVA tele health coaching intervention trial 2005;28(11):2780–6. in Finland; the Greater Green Triangle DPP in Australia; the Peers for Progress 12. Gillies CL, Abrams KR, Lambert PC, Cooper NJ, Sutton AJ, Hsu RT, et al. trial for diabetes self-management in Australia; and the Kerala DPP in India. She Pharmacological and lifestyle interventions to prevent or delay type 2 is also involved in the European Community Horizon 2020 funded SMART2D diabetes in people with impaired glucose tolerance: systematic review and project promoting cross learning in diabetes prevention and management in meta-analysis. Bmj. 2007;334(7588):299. Sweden, South Africa, and Uganda. 13. Orozco LJ, Buchleitner AM, Gimenez-Perez G, Roque IFM, Richter B, Mauricio John Oldroyd PhD is a research fellow in the Department of Epidemiology D. Exercise or exercise and diet for preventing type 2 diabetes mellitus. The and Preventive Medicine, Monash University. His PhD was in lifestyle Cochrane database of systematic reviews. 2008;3:CD003054. interventions for the prevention of diabetes in people with impaired glucose 14. Angermayr L, Melchart D, Linde K. Multifactorial lifestyle interventions in the tolerance in the UK. primary and secondary prevention of cardiovascular disease and type 2 Nicolaas P. Pronk PhD is vice president and chief science officer at diabetes mellitus—a systematic review of randomized controlled trials. HealthPartners and a researcher in the HealthPartners Institute for Education Annals of behavioral medicine: a publication of the Society of Behavioral and Research in Minneapolis, Minnesota, USA. He is also an adjunct professor Medicine. 2010;40(1):49–64. of Social and Behavioral Sciences at the Harvard School of Public Health, 15. Baker MK, Simpson K, Lloyd B, Bauman AE, Singh MA. Behavioral strategies Boston, MA. He undertakes research in the area of health behaviour change in diabetes prevention programs: a systematic review of randomized and population health. He developed the PIPE model that is used in our controlled trials. Diabetes research and clinical practice. 2011;91(1):1–12. systematic review and is an expert in the evaluation of complex 16. Gillett M, Royle P, Snaith A, Scotland G, Poobalan A, Imamura M, et al. Non- interventions. pharmacological interventions to reduce the risk of diabetes in people with Brian Oldenburg PhD is Professor and Chair of Non-Communicable Disease impaired glucose regulation: a systematic review and economic evaluation. Control in the School of Population and Global Health, University of Mel- Health technology assessment. 2012;16(33):1–236. iii-iv. bourne, Australia. He is a behavioural scientist and an expert in the preven- 17. Schellenberg ES, Dryden DM, Vandermeer B, Ha C, Korownyk C. Lifestyle tion and control of chronic conditions. His research program focuses on interventions for patients with and at risk for type 2 diabetes: a systematic improving the prevention and control of cardiometabolic conditions such as review and meta-analysis. Annals of internal medicine. 2013;159(8):543–51. diabetes, heart disease, and their co-morbidities in Australia and other 18. Terranova CO, Brakenridge CL, Lawler SP, Eakin EG, Reeves MM. countries. Effectiveness of lifestyle-based weight loss interventions for adults with type 2 diabetes: a systematic review and meta-analysis. Diabetes Obes Metab. Author details 2015;17(4):371–8. doi: 10.1111/dom.12430. Epub 2015 Jan 14. Melbourne School of Population and Global Health, University of 19. Stevens JW, Khunti K, Harvey R, Johnson M, Preston L, Woods HB, et al. Melbourne, Melbourne 3010, Australia. School of Health Sciences, University Preventing the progression to type 2 diabetes mellitus in adults at high risk: of Tampere, Tampere FI-33014, Finland. Collaborative Care Systems Finland, a systematic review and network meta-analysis of lifestyle, pharmacological Helsinki, Finland. Department of Epidemiology and Preventive Medicine, and surgical interventions. Diabetes Res Clin Pract. 2015;107(3):320–31. doi: Monash University, Melbourne 3004, Australia. HealthPartners Institute for 10.1016/j.diabres.2015.01.027. Epub 2015 Jan 21. Education and Research, 8170 33rd Ave. S, Minneapolis, MN 55425, USA. 20. Yuen A, Sugeng Y, Weiland TJ, Jelinek GA. Lifestyle and medication interventions for the prevention or delay of type 2 diabetes mellitus in Received: 29 April 2015 Accepted: 20 November 2015 prediabetes: a systematic review of randomised controlled trials. Australian and New Zealand journal of public health. 2010;34(2):172–8. 21. Cardona-Morrell M, Rychetnik L, Morrell SL, Espinel PT, Bauman A. Reduction of diabetes risk in routine clinical practice: are physical activity and nutrition References interventions feasible and are the outcomes from reference trials replicable? 1. International Diabetes Federation. IDF diabetes atlas. 6th ed. Brussels: A systematic review and meta-analysis. BMC public health. 2010;10:653. International Diabetes Federation; 2013. http://www.idf.org/diabetesatlas. 22. Taylor J, Cottrell C, Chatterton H, Hill J, Hughes R, Wohlgemuth C, et al. 2. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker Identifying risk and preventing progression to type 2 diabetes in vulnerable EA, et al. Diabetes Prevention Program Research Group. Reduction in the and disadvantaged adults: a pragmatic review. Diabetic medicine : a journal incidence of type 2 diabetes with lifestyle intervention or metformin. The of the British Diabetic Association. 2013;30(1):16–25. New England journal of medicine. 2002;346(6):393–403. 23. Li R, Zhang P, Barker LE, Chowdhury FM, Zhang X. Cost-effectiveness of 3. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, Ilanne- interventions to prevent and control diabetes mellitus: a systematic review. Parikka P, et al. Prevention of type 2 diabetes mellitus by changes in lifestyle Diabetes care. 2010;33(8):1872–94. among subjects with impaired glucose tolerance. The New England journal 24. Jackson L. Translating the diabetes prevention program into practice: a review of medicine. 2001;344(18):1343–50. of community interventions. The Diabetes educator. 2009;35(2):309–20. 4. Pan XR, Li GW, Hu YH, Wang JX, Yang WY, An ZX, et al. Effects of diet and 25. Ali MK, Echouffo-Tcheugui J, Williamson DF. How effective were lifestyle exercise in preventing NIDDM in people with impaired glucose tolerance. interventions in real-world settings that were modeled on the diabetes The Da Qing IGT and diabetes study. Diabetes care. 1997;20(4):537–44. prevention program? Health affairs. 2012;31(1):67–75. 5. Ramachandran A, Snehalatha C, Mary S, Mukesh B, Bhaskar AD, Vijay V, et al. The Indian diabetes prevention programme shows that lifestyle 26. Johnson M, Jones R, Freeman C, Woods HB, Gillett M, Goyder E, et al. Can modification and metformin prevent type 2 diabetes in Asian Indian diabetes prevention programmes be translated effectively into real-world Aziz et al. Implementation Science (2015) 10:172 Page 16 of 17 settings and still deliver improved outcomes? A synthesis of evidence. Diabetic 47. Sepah SC, Jiang L, Peters AL. Translating the diabetes prevention program medicine : a journal of the British Diabetic Association. 2013;30(1):3–15. into an online social network: validation against CDC standards. The 27. Dunkley AJ, Bodicoat DH, Greaves CJ, Russell C, Yates T, Davies MJ, et al. Diabetes educator. 2014;40(4):435–43. Diabetes prevention in the real world: effectiveness of pragmatic lifestyle 48. Kumanyika S, Fassbender J, Phipps E, Tan-Torres S, Localio R, Morales KH, interventions for the prevention of type 2 diabetes and of the impact of et al. Design, recruitment and start up of a primary care weight loss trial adherence to guideline recommendations: a systematic review and meta- targeting African American and Hispanic adults. Contemporary Clinical analysis. Diabetes care. 2014;37(4):922–33. Trials. 2011;32(2):215–24. 28. Pronk NP, Remington PL; Community Preventive Services Task Force. 49. Duijzer G, Haveman-Nies A, Jansen SC, ter Beek J, Hiddink GJ, Feskens EJ. Combined diet and physical activity promotion programs for prevention of Feasibility and potential impact of the adapted SLIM diabetes prevention diabetes: community preventive services task force recommendation intervention in a Dutch real-life setting: the SLIMMER pilot study. Patient statement. Ann Intern Med. 2015; 163. doi:10.7326/M15-1029. education and counseling. 2014;97(1):101–7. 29. Whittemore R. A systematic review of the translational research on the diabetes 50. Kanaya AM, Santoyo-Olsson J, Gregorich S, Grossman M, Moore T, Stewart prevention program. Translational behavioral medicine. 2011;1(3):480–91. AL. The live well, be well study: a community-based, translational lifestyle program to lower diabetes risk factors in ethnic minority and lower- 30. Glasgow RE. Evaluation of theory-based interventions: the RE-AIM model. In: socioeconomic status adults. Am J Public Health. 2012;102(8):1551–8. Glanz K, Lewis FM, Rimer BK, editors. Health behavior and health education. 3rd ed. San Francisco: Jossey-Bass; 2002. p. 531–44. 51. Paulweber B, Valensi P, Lindstrom J, et al. A European evidence-based 31. Laws RA et al. A systematic review of external validity in lifestyle guideline for the prevention of type 2 diabetes. Horm Metab Res. interventions. Am J Prev Med. 2012;43(2):205–14. 2010;42 Suppl 1:S3–36. 32. Green LW, Glasgow RE. Evaluating the relevance, generalization, and 52. National Institute for Health and Clinical Excellence. Preventing type applicability of research—issues in external validation and translation 2 diabetes: risk identification and interventions for individuals at methodology. Eval Health Prof. 2006;29(1):126–53. high risk. London: National Institute for Health and Clinical Excellence; 2012. 33. Pronk NP. Designing and evaluating health promotion programs: simple 53. Balk EM, Earley A, Raman G, Avendano EA, Pittas AG, Remington PL, Task rules for a complex issue. Disease Management & Health Outcomes. Force on Community Preventive Services. Combined diet and physical 2003;11(3):149–57. activity promotion programs to prevent type 2 diabetes among persons at 34. Zyriax BC, Letsch B, Stock S, Windler E. DELIGHT (delay of impaired glucose increased risk: a systematic review for the community preventive services tolerance by a healthy lifestyle trial)—a feasibility study on implementing a task force. Ann Intern Med. 2015;163(6):437–51 doi: 10.7326/M15-0452. program of sustainable diabetes prevention in German companies. Experimental and clinical endocrinology & diabetes: official journal, 54. Li R, Qu S, Zhang P, Chattopadhyay S, Gregg EW, Albright A, Hopkins D, German Society of Endocrinology [and] German Diabetes Association. Pronk NP, [Task Force on Community Preventive Services]. Economic 2014;122(1):20–6. evaluation of combined diet and physical activity promotion programs to 35. Mensink M, Feskens EJ, Saris WH, De Bruin TW, Blaak EE. Study on lifestyle prevent type 2 diabetes among persons at increased risk: a systematic intervention and impaired glucose tolerance Maastricht (SLIM): preliminary review for the community preventive services task force. Ann Intern Med. results after one year. International journal of obesity and related metabolic 2015; 163. doi:10.7326/M15-0469. disorders: journal of the International Association for the Study of Obesity. 55. Lindstrom J, Peltonen M, Eriksson JG, Louheranta A, Fogelholm M, Uusitupa 2003;27(3):377–84. M, et al. High-fibre, low-fat diet predicts long-term weight loss and decreased type 2 diabetes risk: the Finnish diabetes prevention study. 36. Bo S, Ciccone G, Baldi C, Benini L, Dusio F, Forastiere G, et al. Effectiveness Diabetologia. 2006;49:912–20. of a lifestyle intervention on metabolic syndrome. A randomized controlled 56. Yajnik CS, Lubree HG, Rege SS, Naik SS, Deshpande JA, Deshpande SS, et al. trial. J Gen Intern Med. 2007;22(12):1695–703. Adiposity and hyperinsulinemia in Indians are present at birth. The Journal 37. Lakerveld J, Bot S, Chinapaw M, van Tulder M, Kingo L, Nijpels G. Process of clinical endocrinology and metabolism. 2002;87(12):5575–80. evaluation of a lifestyle intervention to prevent diabetes and cardiovascular diseases in primary care. Health promotion practice. 2012;13(5):696–706. 57. Chan JC, Malik V, Jia W, Kadowaki T, Yajnik CS, Yoon KH, et al. Diabetes 38. Ma J, Yank V, Xiao L, Lavori PW, Wilson SR, Rosas LG, et al. Translating the in Asia: epidemiology, risk factors, and pathophysiology. Jama. diabetes prevention program lifestyle intervention for weight loss into primary 2009;301(20):2129–40. care: a randomized trial. JAMA internal medicine. 2013;173(2):113–21. 58. Resnicow K, McMaster F. Motivational interviewing: moving from why to 39. Davis-Smith YM, Boltri JM, Seale JP, Shellenberger S, Blalock T, Tobin B. how with autonomy support. The international journal of behavioral Implementing a diabetes prevention program in a rural African-American nutrition and physical activity. 2012;9:19. church. Journal of the National Medical Association. 2007;99(4):440–6. 59. Spring B, Sohn MW, Locatelli SM, Hadi S, Kahwati L, Weaver FM. Individual, facility, and program factors affecting retention in a national weight 40. Costa B, Barrio F, Cabre JJ, Pinol JL, Cos X, Sole C, et al. Delaying management program. BMC public health. 2014;14:363. progression to type 2 diabetes among high-risk Spanish individuals is 60. Daivadanam M, Absetz P, Sathish T, Thankappan KR, Fisher EB, Philip feasible in real-life primary healthcare settings using intensive lifestyle NE, et al. Lifestyle change in Kerala, India: needs assessment and intervention. Diabetologia. 2012;55(5):1319–28. planning for a community-based diabetes prevention trial. Bmc Public 41. Absetz P, Valve R, Oldenburg B, Heinonen H, Nissinen A, Fogelholm M, et al. Health. 2013;13:95. Type 2 diabetes prevention in the “real world”: one-year results of the GOAL implementation trial. Diabetes Care. 2007;30(10):2465–70. 61. Sathish T, Williams ED, Pasricha N, Absetz P, Lorgelly P, Wolfe R, et al. Cluster 42. Nilsen V, Bakke PS, Gallefoss F. Effects of lifestyle intervention in persons at randomised controlled trial of a peer-led lifestyle intervention program: risk for type 2 diabetes mellitus—results from a randomised, controlled trial. study protocol for the Kerala diabetes prevention program. BMC public BMC public health. 2011;11:893. health. 2013;13:1035. 62. Kosaka K, Noda M, Kuzuya T. Prevention of type 2 diabetes by lifestyle 43. Penn L, Lordon J, Lowry R, Smith W, Mathers JC, Walker M, et al. Translating intervention: a Japanese trial in IGT males. Diabetes research and clinical research evidence to service provision for prevention of type 2 diabetes: practice. 2005;67(2):152–62. development and early outcomes of the ‘new life, new you’ intervention. 63. Oldroyd JC, Unwin NC, White M, Mathers JC, Alberti KG. Randomised British Journal of Diabetes and Vascular Disease. 2011;11(4):175–81. controlled trial evaluating lifestyle interventions in people with impaired 44. Savas LA, Grady K, Cotterill S, Summers L, Boaden R, Gibson JM. Prioritising glucose tolerance. Diabetes Res Clin Pract. 2006;72(2):117–27. prevention: implementation of IGT care call, a telephone based service for people at risk of developing type 2 diabetes. Primary care diabetes. 2015;9(1):3–8. 64. Laatikainen T, Dunbar JA, Chapman A, Kilkkinen A, Vartiainen E, 45. Vanderwood KK, Hall TO, Harwell TS, Butcher MK, Helgerson SD, Montana Heistaro S, et al. Prevention of type 2 diabetes by lifestyle Cardiovascular D, et al. Implementing a state-based cardiovascular disease intervention in an Australian primary health care setting: greater and diabetes prevention program. Diabetes Care. 2010;33(12):2543–5. green triangle (GGT) diabetes prevention project. Bmc Public Health. 2007;7:249. 46. Janus ED, Best JD, Davis-Lameloise N, Philpot B, Hernan A, Bennett CM, 65. Ackermann RT, Finch EA, Brizendine E, Zhou H, Marrero DG. Translating the et al. Scaling-up from an implementation trial to state-wide coverage: diabetes prevention program into the community. The DEPLOY pilot study. results from the preliminary Melbourne diabetes prevention study. Trials. American Journal of Preventive Medicine. 2008;35(4):357–63. 2012;13:152. Aziz et al. Implementation Science (2015) 10:172 Page 17 of 17 66. Boltri JM, Davis-Smith YM, Seale JP, Shellenberger S, Okosun IS, Cornelius ME. Diabetes prevention in a faith-based setting: results of translational research. Journal of Public Health Management and Practice. 2008;14(1):29–32. 67. Payne WR, Walsh KJ, Harvey JT, Livy MF, McKenzie KJ, Donaldson A, et al. Effect of a low-resource-intensive lifestyle modification program incorporating gymnasium-based and home-based resistance training on type 2 diabetes risk in Australian adults. Diabetes Care. 2008;31(12):2244–50. 68. Kramer MK, Kriska AM, Venditti EM, Miller RG, Brooks MM, Burke LE, et al. Translating the diabetes prevention program. A comprehensive model for prevention training and program delivery. American Journal of Preventive Medicine. 2009;37(6):505–11. 69. Kulzer B, Hermanns N, Gorges D, Schwarz P, Haak T. Prevention of diabetes self-management program (PREDIAS): effects on weight, metabolic risk factors, and behavioral outcomes. Diabetes Care. 2009;32(7):1143–6. 70. Penn L, White M, Oldroyd J, Walker M, Alberti KG, Mathers JC. Prevention of type 2 diabetes in adults with impaired glucose tolerance: the European Diabetes Prevention RCT in Newcastle upon Tyne, UK. BMC Public Health. 2009;9:342. 71. Almeida FA, Shetterly S, Smith-Ray RL, Estabrooks PA. Reach and effectiveness of a weight loss intervention in patients with prediabetes in Colorado. Preventing Chronic Disease. 2010;7(5):A103. 72. Makrilakis K, Liatis S, Grammatikou S, Perrea D, Katsilambros N. Implementation and effectiveness of the first community lifestyle intervention programme to prevent type 2 diabetes in Greece. The DE- PLAN study. Diabetic Medicine. 2010;27(4):459–65. 73. Parikh P, Simon EP, Fei K, Looker H, Goytia C, Horowitz CR. Results of a pilot diabetes prevention intervention in East Harlem, New York City: project HEED. American journal of public health. 2010;100 Suppl 1:S232–9. 74. Vermunt PW, Milder IE, Wielaard F, van Oers JA, Westert GP. An active strategy to identify individuals eligible for type 2 diabetes prevention by lifestyle intervention in Dutch primary care: the APHRODITE study. Family practice. 2010;27(3):312–9. 75. Boltri JM, Davis-Smith M, Okosun IS, Seale JP, Foster B. Translation of the National Institutes of Health diabetes prevention program in African American churches. Journal of the National Medical Association. 2011;103(3):194–202. 76. Gilis-Januszewska A, Szybinski Z, Kissimova-Skarbek K, Piwonska-Solska B, Pach D, Topor-Madry R, et al. Prevention of type 2 diabetes by lifestyle intervention in primary health care setting in Poland: diabetes in Europe prevention using lifestyle, physical activity and nutritional intervention (DE-PLAN) project. Br J Diabetes Vasc Dis. 2011;11(4). 77. Katula JA, Vitolins MZ, Rosenberger EL, Blackwell CS, Morgan TM, Lawlor MS, et al. One-year results of a community-based translation of the diabetes prevention program: healthy-living partnerships to prevent diabetes (HELP PD) project. Diabetes Care. 2011;34(7):1451–7. 78. Ruggiero L, Oros S, Choi YK. Community-based translation of the diabetes prevention program's lifestyle intervention in an underserved Latino population. Diabetes Educator. 2011;37(4):564–72. 79. Sakane N, Sato J, Tsushita K, Tsujii S, Kotani K, Tsuzaki K, et al. Prevention of type 2 diabetes in a primary healthcare setting: three-year results of lifestyle intervention in Japanese subjects with impaired glucose tolerance. BMC Public Health. 2011;11(1):40. 80. Ockene IS, Tellez TL, Rosal MC, Reed GW, Mordes J, Merriam PA, et al. Outcomes of a Latino community-based intervention for the prevention of diabetes: the Lawrence Latino diabetes prevention project. American Journal of Public Health. 2012;102(2):336–42. 81. Piatt GA, Seidel MC, Chen H-Y, Powell RO, Zgibor JC. Two-year results of translating the diabetes prevention program into an urban, underserved community. Diabetes Educator. 2012;38(6):798–804. Submit your next manuscript to BioMed Central 82. Jiang LH, Manson SM, Beals J, Henderson WG, Huang HX, Acton KJ, et al. and we will help you at every step: Translating the diabetes prevention program into American Indian and Alaska native communities results from the special diabetes program for • We accept pre-submission inquiries Indians diabetes prevention demonstration project. Diabetes Care. � Our selector tool helps you to find the most relevant journal 2013;36(7):2027–34. � We provide round the clock customer support � Convenient online submission � Thorough peer review � Inclusion in PubMed and all major indexing services � Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Implementation Science Springer Journals

A systematic review of real-world diabetes prevention programs: learnings from the last 15years

Download PDF
17 pages

Loading...

Page 2

Loading...

Page 3

Loading...

Page 4

Loading...

Page 5

Loading...

Page 6

Loading...

Page 7

Loading...

Page 8

Loading...

Page 9

Loading...

Page 10

Loading...

Page 11

Loading...

Page 12

Loading...

Page 13

Loading...

Page 14

Loading...

Page 15

Loading...

Page 16

Loading...

Page 17

 
/lp/springer-journals/a-systematic-review-of-real-world-diabetes-prevention-programs-qmcocEHIsx

References (113)

Publisher
Springer Journals
Copyright
Copyright © 2015 by Aziz et al.
Subject
Medicine & Public Health; Health Promotion and Disease Prevention; Health Administration; Health Informatics; Public Health
eISSN
1748-5908
DOI
10.1186/s13012-015-0354-6
pmid
26670418
Publisher site
See Article on Publisher Site

Abstract

Background: The evidence base for the prevention of type 2 diabetes mellitus (T2DM) has progressed rapidly from efficacy trials to real-world translational studies and practical implementation trials over the last 15 years. However, evidence for the effective implementation and translation of diabetes programs and their population impact needs to be established in ways that are different from measuring program effectiveness. We report the findings of a systematic review that focuses on identifying the critical success factors for implementing diabetes prevention programs in real-world settings. Methods: A systematic review of programs aimed at diabetes prevention was undertaken in order to evaluate their outcomes using the penetration, implementation, participation, and effectiveness (PIPE) impact metric. A search for relevant articles was carried out using PubMed (March 2015) and Web of Science, MEDLINE, CENTRAL, and EMBASE. A quality coding system was developed and included studies were rated independently by three researchers. Results: Thirty eight studies were included in the review. Almost all (92 %) provided details on participation; however, only 18 % reported the coverage of their target population (penetration). Program intensity or implementation—as measured by frequency of contacts during first year and intervention duration—was identified in all of the reported studies, and 84 % of the studies also reported implementation fidelity; however, only 18 % of studies employed quality assurance measures to assess the extent to which the program was delivered as planned. Sixteen and 26 % of studies reported ‘highly’ or ‘moderately’ positive changes (effectiveness) respectively, based on weight loss. Six (16 %) studies reported ‘high’ diabetes risk reduction but ‘low’ to ‘moderate’ weight loss only. Conclusion: Our findings identify that program intensity plays a major role in weight loss outcomes. However, programs that have high uptake—both in terms of good coverage of invitees and their willingness to accept the invitation—can still have considerable impact in lowering diabetes risk in a population, even with a low intensity intervention that only leads to low or moderate weight loss. From a public health perspective, this is an important finding, especially for resource constrained settings. More use of the PIPE framework components will facilitate increased uptake of T2DM prevention programs around the world. Keywords: Implementation, Translational research, Diabetes prevention, Penetration, Implementation, Participation, Effectiveness (PIPE) impact metric, Systematic review, Resource allocation * Correspondence: azizz@student.unimelb.edu.au Melbourne School of Population and Global Health, University of Melbourne, Melbourne 3010, Australia Full list of author information is available at the end of the article © 2015 Aziz et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Aziz et al. Implementation Science (2015) 10:172 Page 2 of 17 Background Another systematic review by Dunkley et al. [27] con- Type 2 diabetes mellitus (T2DM) has emerged as a cluded that pragmatic diabetes prevention programs are major public health challenge with an estimated 387 mil- effective and adherence to international guideline rec- lion people living with T2DM globally [1]. Efforts to ommendations is significantly associated with a greater prevent or delay the onset of diabetes are an urgent weight loss; however, the authors mainly focussed on the public health priority with health, social, and economic effectiveness of the selected translational studies in benefits. Several large randomized controlled trials relation to the main outcome without giving consider- (RCTs) from the US, Finland, China, and India have ation to the adaptability and scalability dimensions of demonstrated that lifestyle interventions can be success- translation. ful in reducing the incidence of T2DM from 29–58 % in A more recent systematic review [28] assessing com- high-risk populations [2–5], with generally good main- bined diet and physical activity promotion programs of tenance for up to 20 years [6, 7]. However, these trials between 3 months and 6 years duration concluded that have mainly focused on efficacy and effectiveness out- more intensive programs resulted in greater weight loss comes. It is now important that more emphasis is given and lowered the risk of T2DM more than less intensive to the implementation and transferability of diabetes programs and this was also found to be cost-effective. prevention programs into real-world settings. ‘Imple- Based on the accumulated evidence from this and other mentation’ research focuses on understanding the pro- reviews, the US Community Preventive Services Task cesses, results, and factors affecting implementation Force (CPSTF) has recently recommended combined under real-world conditions by answering questions diet and physical activity promotion programs to prevent such as ‘why and how interventions work in real-world T2DM [28]. Importantly, however, the above review settings’ [8]. ‘Transferability’ describes the process of ap- examining the translation of diabetes prevention pro- plying the results of research in one situation to other grams like US-DPP did not address issues specifically re- similar situations’ [9]. ‘Real-world’ contexts are settings lated to the long-term scalability and sustainability of where health research findings are applied in practice. such programs. The latter include primary healthcare settings, work or- Another comprehensive review to evaluate the ganizations, churches, and schools. Major transferability generalizability of diabetes prevention trials was con- gaps still remain in translating diabetes prevention from ducted by Whittemore [29] using Glasgow’s[30]reach, ef- research into practice [10]. ficacy, adoption, implementation, maintenance (RE-AIM) Several reviews of diabetes prevention trials have been framework. This is a framework designed to summarize conducted to examine their effectiveness [11–23]. Some how well research trials report on elements related to re- recent reviews have examined how implementation in- search translation. Generalizability is an important elem- fluences effectiveness, most particularly in relation to ent of translational research, i.e. the degree to which the US diabetes prevention program (US-DPP). One of findings from a study or set of studies can be more these reviews [24] summarized lifestyle interventions broadly generalized to populations and settings beyond based on the US-DPP curriculum and described how those in the original studies [31, 32]. This review included different curricula (as a measure of implementation) 16 studies that translated the US-DPP protocol into four affected outcomes. They found that the less-intense distinct settings: (a) hospital outpatient, (b) primary care, version of the US-DPP core curriculum may influence (c) community, and (d) work and church. The author long-term outcomes. Ali et al. [25] conducted a system- found positive outcomes in terms of the reach, efficacy, atic review and meta-analysis of 28 translational studies adoption, implementation, and maintenance of programs. based on the US-DPP and concluded that the utilization However, this review had a narrow focus and only in- of non-medical personnel in delivering diabetes preven- cluded those studies that were based on the US-DPP; no other protocols were considered or included. Another tion interventions can lower the overall costs without compromising effectiveness. Johnson and colleagues limitation was that several studies considered only short- [26] synthesized evidence from translational studies term effectiveness of up to 3 months, and some of the included studies also recruited patients who already had and assessed the impact of interventions delivered outside large randomized trials. The review included diabetes. Laws and colleagues [31] conducted a systematic 17 translational studies from a range of settings and review examining the adequacy of reporting of external concluded that there is potential for less intensive validity components of lifestyle intervention trials interventions to have an impact on future progres- aimed at T2DM prevention. The authors assessed the sion to diabetes in at-risk individuals. However, these generalizability of the findings of 31 studies. Laws re- reviews included studies where subjects had either ported that all studies lacked full reporting on exter- already progressed to diabetes [25, 26] or had short nal validity elements. One of the limitations of this follow-up (e.g. less than 12 months [26]). review was the use of a dichotomous rating scale Aziz et al. Implementation Science (2015) 10:172 Page 3 of 17 (‘reported’ or ‘not reported’) which did not take into Study selection account external validity elements which were re- We included all published studies in the last 15 years ported as continuous measures [31]. (i.e. 2001–2015) that reported on the evaluation of a In summary, a number of recent reviews have focused lifestyle-focused program aimed at individuals at moder- on the real-world effectiveness of T2DM prevention, in ate or high risk of diabetes (e.g. impaired glucose toler- particular, programs derived from the US-DPP. However, ance (IGT), elevated haemoglobin A1c (HbA1c), high these reviews have not systematically examined the body mass index (BMI) or overweight). The inclusion adaptability and scalability dimensions of translation, criteria were adults aged 18 years or older; English lan- that is, what works, under what ‘real-world’ conditions guage publications; and full text. The exclusion criteria and in which contexts; yet this information is critically were studies that were published prior to 2001; that did important for policy makers and program implementers not report at least 1-year follow-up; included partici- who need to identify diabetes prevention programs with pants with known diabetes; included participants not at significant population impact. elevated risk of developing diabetes; and reported on To address this knowledge gap, we have undertaken a multiple intervention components in a single study. systematic review that focuses on identifying the critical Studies were also excluded if they were exclusively ‘diet- success factors for implementing diabetes prevention based’ or ‘exercise-based’ instead of referring to lifestyle programs in real-world settings. We consider the im- or behavioural interventions or if they used lifestyle and portance of program and product design elements that pharmacological interventions. are important to real-world implementation including the effect size, scope of services, scalability potential, Data extraction and long-term sustainability [33]. We use the penetra- An evaluation of T2DM prevention program impact was tion, implementation, participation, effectiveness (PIPE) conducted using the PIPE Impact Metric [33]. The PIPE impact metric, a framework that is highly relevant to im- Impact Metric expresses four elements of program initi- plementation and a formal assessment of the net impact ation and long-term delivery needed to maximize the of health improvement programs that is explicitly linked population impact (i.e., penetration, implementation, to the program design elements noted above [33]. The participation, and effectiveness) as a coefficient. The four key elements of the PIPE framework are as follows: product of all coefficients becomes the PIPE Impact (1) penetration of the program into the population of Metric. For example, a study may report having reached interest, (2) implementation of the proposed set of ser- 250 out of 350 individuals at high risk for T2DM vices, (3) participation in the program, and (4) effective- workers at a company. In this example, penetration is re- ness in generating expected outcomes. Each of the PIPE ported as (250/350) = 0.714 (71 %). Similar coefficients Impact Metric elements may be expressed as a coeffi- are calculated for implementation, participation, and ef- cient, and the product of these four coefficients is re- fectiveness. As numeric data was not available for some ferred to as the PIPE impact metric. The PIPE Impact of the elements in some of the included studies, an alter- Metric can be used to provide feedback to program native coding system, also informed by PIPE, was devel- administrators on gaps in performance, and it also oped for this paper. enables the integration of program design features and identifies where to focus on for performance improve- Data synthesis and analysis ment changes. This paper aims to review the current The coding system is summarized in Table 1. It included evidence about success factors for implementing dia- two steps: (1) an initial scoring and (2) coding of the betes prevention programs in real-world settings using scores into ‘high’, ‘medium’, ‘low’,or ‘not able to calculate the PIPE Impact Metric. (NAC)’ where relevant data was not available. In the ini- tial scoring, coefficients were calculated for penetration and participation. For implementation, the initial scoring Methods was done on three aspects: frequency over the first Data sources and searches 12 months; duration of the entire intervention including A comprehensive search was carried out using PubMed, follow-up contacts; and intervention fidelity. To over- Web of Science, MEDLINE, CENTRAL, and EMBASE come heterogeneity in the kinds of contacts when scor- (February 2014). Search terms were ‘diabetes’ AND ‘pre- ing frequency, a system was developed to standardize vention’ AND (‘program’ OR ‘intervention’) AND (im- the degree of contact based on number, length, and type plementation’ OR ‘translation’). The search was repeated (Table 1). Similarly, for effectiveness, the initial scoring using PubMed to include relevant articles from February was on three criteria: proportion of participants successful 2014 to March 2015. A detailed search strategy is pro- in achieving the main outcome; weight loss (in kilo- vided in Additional file 1. grammes); and diabetes risk reduction (absolute/relative). Aziz et al. Implementation Science (2015) 10:172 Page 4 of 17 Table 1 Development of coding system using PIPE Impact Metric elements PIPE Element Description Coding of the scores Penetration Numerator: the number of individuals reached (invited) ≤33 % = low; 34–66 % = moderate; ≥67 % = high; or NAC (not able to calculate) Denominator: the number of individuals in target population Implementation Implementation was rated on 3 aspects. ≤33 % = low; 34–66 % = moderate; or ≥67 % = high 1. Frequency: the degree of contact (based on number, length, and type) over the first 12 months of an intervention. Different types of contacts were quantified based on the session type in the following way: � 1 group/individual session = 1 session � 1 group/individual session (>3 h) = 2 sessions � 1 online/telephone session = 0.5 session � 1 text/email/fax contact = 0.25 session Numerator: total number of sessions (over the first 12 months) Denominator: 22 (the US-DPP 16 weekly + 6 monthly = 22 sessions) 2. Duration: the duration of the intervention ≤6 months = low; 6–12 months = moderate; and >12 months = high 3. Fidelity: the use of standard curriculum (for example: the No standard curriculum followed = low; a standard curriculum US-DPP) for the delivery of intervention and use of quality was followed but no quality assurance measures were assurance measures to monitor the implementation of the reported = moderate; a standard curriculum was followed intervention and quality assurance measures were applied = high; or NAC (not able to calculate) Participation Numerator: the number of participants enrolled in the intervention ≤33 % = low; 34–66 % = moderate; ≥67 % = high; or NAC (not able to calculate) Denominator: the number of individuals reached (invited) Effectiveness Effectiveness was rated on 3 criteria: ≤25 % = low; 26–40 % = moderate; >40 % = high; or NAC (where information is not provided) 1. Success criterion/proportion of successful participants: Numerator: participants who achieved the main outcome (i.e. weight loss ≥5%) Denominator: total number of participants enrolled in the intervention/total number of participants completed 12-month measurements 2. Average weight loss: the average weight loss (in kilogrammes) ≤2.3 kg = low; 2.4–4.6 kg = moderate; >4.6 kg = high; or NAC (where information is not provided) 3. Risk reduction: diabetes risk reduction (absolute/relative) Risk reduction: ≤15 % = low; 16–30 % = moderate; >30 % = high; or NAC (where information is not provided) The coding of the scores (into high, medium, low, and assessed for eligibility. A total of 76 articles from 38 NAC) for each element was agreed-upon by the review studies were included in the review. Table 2 describes team, and ratings were independently derived by three re- the characteristics of the included studies. searchers (ZA, PA, JO). The ratings of all studies were Studies that met the eligibility criteria for this review then reviewed by all authors, and when disagreement on were mainly based on either the US-DPP or Finnish-DPS. rating occurred, they were resolved with consensus among Seventeen (45 %) studies were implemented in the USA, 4 all authors. in the UK, 4 in the Netherlands, 8 in other European countries, 3 in Australia, and 2 in Japan. There were no Results studies from low and middle income countries that met A detailed PRISMA flow diagram is attached (Fig. 1). our eligibility criteria. The sample size for the participants The initial literature search (February 2014) returned enrolled in each of these studies ranged between 8 and 2992 publications and 61 additional articles were identi- 2553 participants. The studies were conducted in a range fied through hand searching of references from the bibli- of settings including primary health care, faith-based, ographies of articles identified. Two thousand thirty- workplace, and other community-based settings. nine articles were screened after removing duplicates. Each study was assessed on the four components of An additional 5 articles were included after updating the PIPE Impact Metric using a coding system described search until March 2015. A total of 180 articles were in Table 1. Table 3 describes the ratings of all Aziz et al. Implementation Science (2015) 10:172 Page 5 of 17 Records identified Additional articles through database identified through searching hand searching (n = 2,992) (n=61) Records after duplicates removed (n = 2,039) Records excluded by Records screened title/abstract (n = 2,039) (n = 1,864) Additional studies (from February 2014 to March 2015) (n = 5) Full-text articles excluded, with reasons Full-text articles n = 104 (92 studies) assessed for eligibility Reasons: (n = 180) * < 12-months follow-up (n = 19) * Included people with T2DM (n = 19) * Study protocol/design (n = 13) * Included young/healthy people (n = 9) * Included multiple interventions (n = 7) * No full article found (n = 7) Full-text articles * Published prior to 2001 (n = 3) included in this review * Other (n = 15) n = 76 articles (38 studies) Fig. 1 PRISMA flow diagram included studies based on the elements of the PIPE only study rated as having ‘moderate’ penetration, was a Impact Metric. church-based study [39]. The church roster included 407 members, whereas 37 % of adults (approximately Program penetration 150) who attended Sunday gatherings were invited to Our analysis shows that only nine (24 %) studies re- complete a diabetes risk assessment. The study rated as ported their estimated target population, from which having ‘low’ penetration in our analysis [40] included 18 only seven studies reported the proportion of the target participating centres that covered all primary care ser- population that was reached with invitations to engage vices for 4.5 % of the population in Catalonia, out of in the program or intervention. Out of these seven stud- which less than 1 % of the population was invited for ies, five had ‘high’, one had ‘moderate’, and one had ‘low’ screening. penetration into their target populations (see Table 3 and Additional file 2: Table S4). Target populations in- Program implementation cluded patients, employees, and church attendees. Each In order to assess implementation, we evaluated the de- study used various strategies to recruit potential partici- gree of contact (based on number, length, and type) dur- pants including mail invitations, posted flyers, advertis- ing the first year of the intervention as frequency; the ing through media, contacting local physicians, local duration of the entire intervention; and the fidelity of churches, or using intranet or work meetings in the the intervention (see Table 1). All studies in our analysis workplace setting. reported on frequency and duration. Thirty-four percent The five studies that were rated as having ‘high’ pene- of all studies implemented ‘high’ frequency interven- tration in our analysis applied heterogeneous strategies tions, and 39 % studies delivered intervention over the to reach their target group. Two studies contacted a pre- period of 12 months or more. The number of contacts defined group of people at risk: one at worksite, where varied from a single small group session to 32 group ses- all employees who had above average waist circumfer- sions. About two thirds (66 %) of the programs based on ence were invited for screening [34], and in the other the US-DPP model adopted a ‘low-’ to ‘moderately in- study [35] all eligible subjects with high risk for glucose tense’ version (based on the degree of contact) as com- intolerance from a cohort representing general popula- pared to the original. The main adaptation was the tion were contacted. Three studies contacted 100 % [36], reduction from 16 to fewer sessions. In addition, groups 68 % [37], and 70 % [38] of the target population for se- led by volunteers as opposed to health care professionals lective screening either by mail or by appointment. The and use of telephone as opposed to face-to-face delivery Included Eligibility Screening Identification Aziz et al. Implementation Science (2015) 10:172 Page 6 of 17 Table 2 Characteristics of the included studies Year Author Study ID Country Setting Study population Sample Study design Intervention size 2003 Mensink et al. SLIM Netherlands Unclear Adults at risk of T2DM 114 RCT 3 individual and 1 group session during 1 year + [35] participants were encouraged to participate in the exercise program 3 times a year 2005 Kosaka et al. Japanese DPP Japan Hospital-based Adults with IGT 458 RCT Detailed instructions on lifestyle were repeated every [62] 3 to 4 months during hospital visits 2006 Oldroyd et al. Newcastle lifestyle UK Primary care Adults with IGT 78 RCT 12 individual 15- to 20-min review appointments over [63] intervention 24 months (6 in the first 6 months, 1 after 9 months and 5 at 2 monthly intervals between 12 and 24 months) 2007 Absetz et al. GOAL LIT Finland Primary care Adults at risk of T2DM 352 Before and after Six 2-hourly group counselling sessions over 8 months [41] 2007 Bo et al. [36] Italian Trial Italy Primary care Adults with metabolic syndrome 335 RCT 1 individual and four 1-hourly group sessions 2007 Davis-Smith DPP (church- USA Community Adults at risk of T2DM 10 Before and after 6-session lifestyle intervention delivered over a 7 week et al. [39] based) (church) period 2007 Laatikainen Greater Green Australia Primary care Adults at risk of T2DM 311 Before and after 6 structured 90-min group sessions delivered during et al. [64] Triangle (GGT) an 8-month period 2008 Ackermann DEPLOY USA Community Adults at risk of T2DM 92 RCT Sixteen 1 to 1.5-hourly small group sessions over 16 to et al. [65] (YMCA) 20 weeks and monthly large-group meetings 2008 Boltri et al. DPP (church- USA Community Adults with pre-T2DM 8 Before and after 16 weekly group sessions conducted over 4 months [66] based) (church) 2008 Payne et al. BDPPI Australia Outpatient setting Adults at risk of T2DM 122 Before and after 6-week group self-management education program, [67] 12-week gym- or home-based resistance training, and three 2-h group reinforcement sessions during 34-week maintenance program 2009 Kramer et al. GLB (2007–2009) USA Community Adults with pre-T2DM 42 Before and after 12 weekly sessions (~60 min) and participants were [68] offered the opportunity to attend monthly support meetings for 9 months after completion of the intervention 2009 Kulzer et al. PREDIAS Germany Outpatient setting Adults at risk of T2DM 182 RCT 12 lessons lasting ~90 min each [69] 2009 Penn et al. EDIPS—Newcastle UK Outpatient setting Adults with IGT 102 RCT A 30-min session immediately following randomisation [70] and 2 weeks later, then monthly for the first 3 months and every 3 months thereafter up to 5 years 2010 Almeida et al. Colorado weight USA Integrated health Adults with pre- T2DM 1520 Matched cohort A single 90-min small group session [71] loss intervention care organization 2010 Makrilakis et al. DE-PLAN Greece Greece Primary care Adults at risk of T2DM 191 Before and after 6 sessions (1 h each) held by a registered dietician in [72] (workplace) the groups of 6 to 10 persons 2010 Parikh et al. Project HEED USA Community Adults at risk of T2DM 99 RCT A peer-led lifestyle intervention group, presented in a [73] workshop consisting of eight 1.5-h sessions over 10 weeks Aziz et al. Implementation Science (2015) 10:172 Page 7 of 17 Table 2 Characteristics of the included studies (Continued) 2010 Vanderwood Montana CVD and USA Health care Adults at risk of T2DM and CVD 355 Before and after 16 weekly group sessions and 6 monthly group et al. [45] DPP facilities (pilot study) sessions 2010 Vermunt et al. APHRODITE Netherlands Primary care Adults at risk of T2DM 925 RCT 11 consultations of 20-min over 2.5 years, five 1-h [74] group meetings and 1-h personal consultation with the dietician 2011 Boltri et al. DPP (church- USA Community Adults with pre-T2DM 37 Before and after 6 or 16 weekly group sessions [75] based) (church) 2011 Gilis- DE-PLAN Poland Poland Primary care Adults at risk of T2DM 175 Prospective cohort 10 group sessions lasting for 4 months, 6 telephone Januszewska motivation sessions, and 2 motivation letters sent to et al. [76] participant 2011 Katula et al. HELP PD USA Community Adults with pre-T2DM 301 RCT ~26 weekly group sessions for the first 6 months, 3 [77] (various venues) personalized consultations with a registered dietician, 18 monthly group sessions, and monthly phone contact for the last 18 months 2011 Kumanyika Think health! USA Primary care Adults with high BMI and 261 RCT Brief monthly contact with a lifestyle coach (LC) for et al. [48] weight 12 months and 10–15 min counselling sessions with primary care providers every 4 months. Bi-monthly sessions with LC for the second year 2011 Nilsen et al. Nilsen et al. Norway Primary care Adults at risk of T2DM 213 RCT The individual and interdisciplinary group participated [42] in a group-based program, 1 day (5 h per day) each week for 6 weeks An individual 30-min consultation with a nurse or ergonomist completed the intervention 1 month after the last group meeting 2011 Penn et al. NLNY UK Leisure and Adults at risk of T2DM 218 Before and after A 10-week program of twice-weekly 1.5-h sessions, [43] community followed by ongoing support with regular mobile settings phone text message and email reminders, ‘drop-in’ activity sessions continued up to 12 months 2011 Ruggiero et al. HLP USA Community Adults at risk of T2DM 69 Before and after 16 weekly core sessions and 6 monthly after-core [78] (various venues) sessions 2011 Sakane et al. Japanese Study Japan Primary care Adults with IGT 304 RCT 4 group sessions of 2 to 3 h (for the first 6 months), [79] (workplace) individual sessions twice a year for 3 years. Between-visit contact by fax was also made monthly during the initial 12 months 2012 Costa et al. DE-PLAN-CAT Spain Primary care Adults at risk of T2DM 552 Prospective cohort A 6-h educational program (scheduled in 2 to 4 [40] individual/small group sessions), and regular contact by phone or text message for at least once every 6 to 8 weeks 2012 Janus et al. pMDPS Australia Primary care Adults at risk of T2DM 92 RCT 6 structured 90-min group sessions. The first 5 sessions [46] were at 2 weeks intervals and the final session was 8 months after the first 2012 Kanaya et al. LWBW USA Community Adults at risk of T2DM 238 RCT The intervention was primarily telephone-based [50] counselling (12 calls) with 2 in-person sessions Aziz et al. Implementation Science (2015) 10:172 Page 8 of 17 Table 2 Characteristics of the included studies (Continued) and 5 optional group workshops over 1 year period 2012 Lakerveld et al. Hoorn Prevention Netherlands Primary care Adults at risk of T2DM and/or 622 RCT Six individual 30-min counselling sessions, [37] Study CVD followed by 3-monthly booster sessions by phone for a period of 1 year. 2012 Ockene et al. LLDPP USA Community Adults at risk of T2DM 312 RCT 3 individual and 13 group sessions over a [80] 12 month period 2012 Piatt et al. [81] GLB (2005–2008) USA Community Adults with metabolic syndrome 105 Before and after 12 weekly sessions over 12 to 14 weeks (lasted ~90 min) in the groups of 5 to 13 participants 2013 Jiang et al. SDPI-DP USA Community Adults with pre-T2DM 2553 Before and after 16 group sessions in the first 16 to 24 [82] weeks and monthly individual lifestyle coaching sessions 2013 Ma J et al. [38] E-LITE USA Primary care Overweight/obese adults with 241 RCT 12 weekly group sessions (1.5 to 2 h each) in increased cardiometabolic risk the first 3 months. From month 4 to 15, contact every 2 to 4 weeks depending on participant needs and preferences. Individual, secure email/ phone contacts with personalized progress feedback and lifestyle coaching throughout the maintenance phase (month 4 to 15) 2014 Duijzer et al. SLIMMER Netherlands Primary care Adults at risk of T2DM 31 One group pre-test In addition to 6 individual consultations [49] post-test (in total 4 h per participant), on average, participants received 5.2 consultations by dieticians and 34.1 sports lessons 2014 Sepah et al. Prevent USA Online platform Adults with pre-T2DM 220 Quasi-experimental 16 online weekly lessons. Participants were then [47] research design offered to continue with a post-core lifestyle change maintenance intervention, with the entire intervention (core plus post-core) totalling 12 months 2014 Zyriax et al. DELIGHT Germany Primary care Adults at risk of T2DM 241 Before and after 12 weekly sessions (for the first 6 months), 6 monthly [34] (workplace) and 6 biweekly sessions (for the next 6 months). For year 2 and 3 quarterly 1.5-h sessions 2015 Savas et al. IGT care call UK Primary care Individuals with IGT 55 Observational A telephone service providing a 6 month lifestyle [44] study education program (20 min × 6), in addition to an introduction call (10 min) and action planning call (40 min) Aziz et al. Implementation Science (2015) 10:172 Page 9 of 17 Table 3 Scoring for each PIPE element by study Author Year Study Penetration Implementation Participation Effectiveness Frequency Duration Fidelity Success rate Weight loss Risk reduction (absolute/relative) Mensink et al. 2003 [35] SLIM High Low High Low Low NAC Moderate High Kosaka et al. 2005 [62] Japanese DPP NAC Low High Moderate NAC NAC Moderate High Oldroyd et al. 2006 [63] Newcastle LI NAC Moderate High NAC Low NAC Low NR Absetz et al. 2007 [41] GOAL LIT NAC Moderate Moderate Moderate High NAC Low NR Bo et al. 2007 [36] Italian Trial High Low High High Low NAC Low High Davis-Smith et al. 2007 [39] DPP (church-based) Moderate Low Low Moderate Low NAC High NR Laatikainen et al. 2007 [64] GGT NAC Low Moderate High Low NAC Moderate NR Ackermann et al. 2008 [65] DEPLOY NAC High High Moderate Low NAC High NR Boltri et al. 2008 [66] DPP (church-based) NAC High Low Moderate Low NAC Low NR Payne et al. 2008 [67] BDPPI NAC High High Moderate NAC Moderate Moderate NR Kramer et al. 2009 [68] GLB (2007 – 2009) NAC High Moderate Moderate Low NAC High NR Kulzer et al. 2009 [69] PREDIAS NAC Moderate Low Moderate NAC NAC Moderate NR Penn et al. 2009 [70] EDIPS- Newcastle NAC Moderate High NAC Low NAC Low High Almeida et al. 2010 [71] Colorado weight loss intervention NAC Low Low NAC Low Low Low NR Makrilakis et al. 2010 [72] DE-PLAN Greece NAC Low Moderate Low Low NAC Low NR Parikh et al. 2010 [73] Project HEED NAC Moderate Low Moderate Low Moderate Moderate NR Vanderwood et al. 2010 [45] Montana CDDP NAC High Moderate Moderate High High High NR Vermunt et al. 2010 [74] APHRODITE NAC Moderate High NAC Low NAC NAC NR Boltri et al. 2011 [75] DPP (church-based) NAC Low (2 churches) Low Moderate Low NAC Low NR High (3 churches) Gilis-Januszewska et al. 2011 [76] DE-PLAN Poland NAC Moderate Moderate Low Low Low Low NR Katula et al. 2011 [77] HELP PD NAC High High High Low NAC High NR Kumanyika et al. 2011 [48] Think Health! NAC Moderate Moderate Moderate Moderate Low Low NR Nilsen et al. 2011 [42] Nilsen et al. NAC High High NAC High Moderate NAC NR Penn et al. 2011 [43] NLNY NAC High Moderate Low High Low Low NR Ruggiero et al. 2011 [78] HLP NAC High Moderate Moderate Low Moderate Low NR Sakane et al. 2011 [79] Japanese Study NAC Moderate High Moderate Low NAC Low High Costa et al. 2012 [40] DE-PLAN-CAT Low Low High NAC Low NAC Low High Janus et al. 2012 [46] pMDPS NAC Low Moderate High High NAC Moderate Moderate Kanaya et al. 2012 [50] LWBW NAC Moderate Moderate Moderate Moderate NAC Low NR Lakerveld et al. 2012 [37] Hoorn Prevention Study High Moderate Moderate High Low NAC NAC NR Aziz et al. Implementation Science (2015) 10:172 Page 10 of 17 Table 3 Scoring for each PIPE element by study (Continued) Ockene et al. 2012 [80] LLDPP NAC High Moderate Moderate Low NAC Low NR Piatt et al. 2012 [81] GLB (2005–2008) NAC Moderate Moderate Moderate Low Moderate NAC NR Jiang et al. 2013 [82] SDPI-DP NAC High Moderate Moderate Low NAC Moderate NR Ma J et al. 2013 [38] E-LITE High High High High Low High High NR Duijzer et al. 2014 [49] SLIMMER NAC Low Moderate Moderate Moderate NAC Moderate NR Sepah et al. 2014 [47] Prevent NAC Moderate High Moderate High High NAC NR Zyriax et al. 2014 [34] DELIGHT High High High Low Low NAC NAC NR Savas et al. 2015 [44] IGT Care Call NAC Low Moderate High High NAC Moderate NR Details on the scoring of all included studies based on the elements of the PIPE Impact Metric framework are provided in Additional file 2: Table S4–Table S7 NAC not able to calculate, NR not reported Aziz et al. Implementation Science (2015) 10:172 Page 11 of 17 of individual sessions were also frequently observed ad- Also, none of the seven studies that were rated as ‘high’ aptations. Only a small proportion (16 %) of studies re- participation provided enough information on reaching ported ‘low’ duration, i.e. intervention delivered over the out to their target populations, and hence, penetration period of 6 months or less. could not be calculated. However, the information avail- Implementation fidelity was defined as intervention able suggests that the studies where high-risk partici- being based on a standard curriculum for example the pants were identified and referred through their GPs or US-DPP and whether any quality assurance measures nurses resulted in ‘high’ participation rates. were applied to monitor the implementation of the intervention. Twenty-seven (71 %) studies were based Program effectiveness on a standard curriculum, out of which only 7 studies Effectiveness was rated based on three criteria: propor- had quality assurance measures applied to monitor the tion of successful participants; average weight loss; and implementation. In some of these studies, while authors diabetes risk reduction (absolute/relative) (Refer to reported on the efforts made to minimize the potential Table 1). None of the studies reported on all three cri- lack of fidelity, none of these studies provided informa- teria. Seventeen (45 %) studies reported the use of tion on the extent to which the various components intent-to-treat analysis; however, for the purpose of our were delivered except one where the program compo- analysis, effectiveness indicators were considered as pre- nents were more frequently added (40 %) than omitted sented in each of the studies. (28 %) [41]. Refer to Table 3 and Additional file 2: Table One third (n = 12, 32 %) of the studies reported the S5 for further details. proportion of successful participants who achieved the primary outcome (i.e. 5 % weight loss). The proportion Program participation of successful participants ranged between 20 and 64 %. The majority of the studies in our analysis (n = 35; 92 %) Thirty-two (84 %) studies reported average weight loss reported participation. Twenty-five (71 %) of these 35 by participants at 12 months, with a range from 0.45 to studies achieved ‘low’ participation rates. Half of these 7.7 kg. Only six of these studies were rated ‘high’ where studies achieved participation rates equivalent to or average weight loss by participants was more than lower than 10 %. Only 7 (18 %) studies had ‘high’ partici- 4.6 kg. Sixteen (42 %) studies were rated ‘low’ on the pation rates. For four of these studies [41–44], partici- basis of average weight loss of ≤2.3 kg. Only seven pants were recruited by referral from physicians, general (18 %) studies reported the data on diabetes risk re- practitioners (GPs), or nurses from the participating duction (absolute/relative), where six studies were health facilities and invited to attend a screening clinic. rated ‘high’ and one was rated ‘moderate’.Scores Two of the 7 studies [45, 46] used a combination of could not be calculated in three studies due to the strategies to recruit potential participants including con- lack of numerical data. tacting local physicians and primary healthcare practices; Thirteen studies (34 %) were rated ‘high’ on at least advertising through media; and recruiting through local one of the three criteria. One study [45] that was rated employers, work sites, churches, and service groups. ‘high’ on both success rate and weight loss, reported that One study [47] recruited participants from online ad- 64 % of participants achieved 5 % weight loss goal with vertisement, seeking individuals with a self-reported the average 12-month weight loss of 7.7 kg. This study clinical diagnosis of prediabetes occurring within the was not only ‘highly’ effective but also had ‘high’ partici- past year; however, recruitment was based on self- pation rates. The intervention included ‘high’ number of selection by participants, which does not reflect a sessions with ‘moderate’ duration and ‘moderate’ fidelity. truly random sample [47]. Another study [38] that had ‘high’ penetration into its Three studies were scored as having ‘moderate’ partici- target population appeared to be ‘highly’ effective based pation, where 63 % [48], 57 % [49], and 44 % [50] of the on the proportion of successful participants (50 %) and individuals were enrolled after assessing all ‘invited’ indi- average weight loss (6.3 kg at 15 months). This study viduals, for study eligibility (see Table 3 and Additional however, reported ‘low’ participation and delivered file 2: Table S6). In most of the remaining studies, ‘high’ number of sessions and had ‘high’ duration of the ‘low’ participation was attributed to either the intervention. non-eligibility of potential participants or the refusal Five out of six studies that reported ‘high’ weight loss to participate. delivered ‘high’ number of sessions. For studies where information was available for both Out of 10 studies that reported ‘moderate’ weight loss, penetration and participation, it was observed that ‘high’ 6 (60 %) implemented interventions with ‘low’ number penetration into the target population did not have posi- of sessions/contacts. Sixteen (42 %) studies had ‘low’ ef- tive effect on participation. All five studies rated as ‘high’ fectiveness based on average weight loss, out of these, 8 penetration in the analysis, reported ‘low’ participation. (50 %) studies delivered ‘moderate’ number of sessions, Aziz et al. Implementation Science (2015) 10:172 Page 12 of 17 whereas 4 (25 %) delivered ‘high’ and 4 (25 %) delivered ‘high’ penetration resulted in either ‘high’ weight loss ‘low’ number of sessions. Six studies reported ‘high’ ef- or ‘high’ diabetes risk reduction. Hence, scalability of fectiveness in risk reduction despite ‘low’ (4 studies) or the program to reach a large audience appears to be ‘moderate’ (2 studies) effectiveness in weight loss, and an important ingredient for population-level impact. these were also all studies with only ‘low’ (4 studies) or In summary, while an intensive intervention plays an ‘moderate’ (2 studies) frequency, but with ‘high’ duration important role in achieving successful weight loss out- (6 studies). See Table 3 and Additional file 2: Table S7 comes, highly scalable moderate- to low-frequency inter- for more details. ventions appear to have major potential to achieve diabetes risk reduction in populations. Discussion From a program implementation perspective, it is im- This is the first systematic review to evaluate the imple- portant to clearly define and estimate the size of the tar- mentation of real-world diabetes prevention programs get population. Without a reasonable estimate of the using the PIPE Impact Metric framework that deploys size of the target population, there is a risk that the pro- four highly relevant elements for monitoring program gram will not be scalable or sustainable [33]. We found impact in real-world settings. As such, this review com- estimates of target populations to be reported by less plements other recent reviews, e.g. Dunkley et al. 2014 than one third of included studies. Hence, it is possible [27], by providing a more detailed understanding of key that many of these programs were delivered to highly se- factors underlying successful translation and implemen- lected populations which limit their generalizability. tation of diabetes prevention programs in real-world We used the US-DPP intervention as a benchmark to contexts. We have also defined the specific scope of assess the intensity of dose-delivered (including fre- services for calculating the overall costs of services being quency, duration, and fidelity) in our included studies. provided. From both an organizational and societal In order to deal with heterogeneity in the types of con- perspectives, these issues are important to consider since tacts, we constructed a framework to standardize the the relative costs and benefits of such services and degree of contact based on number, length, and type. programs are important determinants of their uptake We found that many programs adopted a ‘low-’ to ‘mod- and adoption. Our review of studies published over erately intense’ version, compared to the US-DPP inter- the last 15 years aims to identify the components of vention. In practice, this also means that many of the diabetes prevention programs with the highest popu- components were either omitted or modified from the lation impact. original US-DPP curriculum. As stated earlier, we also Our review highlights several important findings. First, note that ‘high’-frequency interventions with ‘high’ to confirming earlier reviews, our analysis demonstrates ‘moderate’ duration and fidelity were associated with that lifestyle-focused diabetes prevention programs that greater weight loss. This is consistent with a systematic have a ‘high’ degree of contact have more potential to review and meta-analysis by Dunkley et al. [27], where achieve effective outcomes, especially when measured by reviewers coded intervention content based on the weight loss. These programs have typically been based recommendations for lifestyle interventions for the pre- on the US-DPP model and have used a very structured vention of diabetes provided by both the European protocol to maximize program fidelity. However, the Guideline and Training Standard for Diabetes Preven- problem with this approach is that in these studies, tion (IMAGE) project [51] and National Institute for program participation tends to be quite low; and fur- Health and Care Excellence (NICE) [52] and found that thermore, none of these studies reported diabetes risk adherence to guidelines on the content and delivery was reduction. significantly associated with a greater weight loss. Our Second, six of the studies showed different degrees of findings on the potential of ‘low’ to ‘moderate’ frequency effectiveness for different outcomes. For example, dia- interventions with longer duration to achieve significant betes risk reduction could be ‘high’ even when effective- risk reduction support earlier findings [25, 26]. A recent ness in weight loss was ‘low’ or ‘moderate’. Surprisingly, CPSTF review [28] shows lower weight loss than the these were all studies of ‘low’ or ‘moderate’ frequency, US-DPP but still concluded strong evidence of effective- but ‘high’ duration. This could be very promising espe- ness. The recommendations from this review were fur- cially for settings where intervention resources are con- ther supported by an effectiveness and economic review strained but when large populations can be reached by [53, 54]. So, in general, it seems that studies beyond the such programs. original US-DPP generate somewhat lower effect for Third, we found that ‘high’ penetration into the tar- weight loss but still generate meaningful positive impact get population with invitations to engage prospective on the reduction of incidence of T2DM. Lindstrom et al. participants in the program do not necessarily result have also previously discussed the positive effects of de- in ‘high’ participation. However, three studies with creased fat and increased fibre intake on diabetes risk Aziz et al. Implementation Science (2015) 10:172 Page 13 of 17 reduction in the absence of weight loss [55]. We also used either criterion—diabetes risk reduction or weight know that in many populations diabetes risk can be loss—to assess effectiveness. high at lower levels of weight, with other factors be- We found that weight loss was reported in several dif- sides weight loss playing a critical role in risk reduc- ferent ways. For example, the percentage weight lost tion [56, 57]. could be interpreted to mean weight loss to percentage Examining the implementation component further, in achieving a particular weight loss target or the average calculating ‘frequency’, we have used contacts made in percentage of weight that subject lost. One study [45] re- the initial 12 months only because most of the studies ported both 5 and 7 % weight loss, whereas another [40] did not extend beyond 12 months. In those that did, the reported the proportion of successful participants who initial 12 months can be considered as the ‘action’ phase, achieved at least 3 % weight loss. For consistency, in our bringing about the lifestyle changes, and beyond that is a review, we only used 5 % weight loss when it was re- follow-up and maintenance phase, which some studies ported in several different ways. support with less frequent contacts. However, examination of studies with ‘low’ effective- In translational research, a systematic evaluation of ness reveals that for some of the studies the reported program fidelity is important to assess the extent to changes in weight loss were very small. Some of these which program was implemented as designed. We based studies reported a significant reduction in weight follow- our definition of fidelity on whether a standard curricu- ing the active intervention phase, but the weight was lum was used to guide the delivery of intervention and partly or, in some studies, entirely regained by the end of whether quality assurance measures were placed to 12 months. Lack of consistency in the way weight loss monitor the implementation of the intervention. Not outcomes are reported and analysed needs to be ad- many studies clearly reported whether the quality as- dressed in future translational research [29]. Cardona- surancemeasureswereabletoguarantee ‘high’ fidel- Morrell and colleagues [21] suggests the establishment ity, this clearly being one of the next important steps of a registry of translational projects using consistent, in program development. This needs to be examined measurable outcomes to add more certainty to effective- in future studies. ness analyses. All but three studies in our review reported participa- Our review includes diabetes prevention translational tion, and only seven studies scored ‘high’ on participa- programs published since 2001 and until 2015. The tion rates. The high participation may be a reflection of studies included in this review have implemented ‘high’ highly targeted penetration, but because of unavailability (34 %), ‘moderate’ (37 %), and ‘low’ (29 %) frequency in- of information, we could not calculate penetration coef- terventions; however, we noticed that most of the ‘low- ficients for any of these seven studies. However, the in- frequency’ interventions were conducted in earlier years, formation available suggests that the studies where high- whereas, designing ‘moderate’ to ‘high’ intensity inter- risk participants were identified and referred through ventions occurred in more recent years. However, ‘par- their GPs or nurses resulted in ‘high’ participation rates, ticipation’ has been consistently low in a majority of the underlying the important role of providers. The high studies over the last 15 years. One of the reasons for this participation could also be a reflection of high motiv- may be the fact that program planners focus on the con- ation for change among at-risk individuals, a factor that tent of the interventions instead of balancing the con- the providers can further enhance in face-to-face con- tent with the experience of the participant—that is, on tact, e.g. by applying strategies from motivational inter- the ‘participation’ dimension and the engagement factor viewing [58]. In our review, of the seven studies that that connects the participant with the intervention. scored ‘high’ for participation, five studies reported ‘high’ Future translational research in this field needs to in- to ‘moderate’ effectiveness based on success rate and/or vest in designing recruitment more carefully to ensure weight loss. high program reach; examining factors that optimize The PIPE Impact Metric elements are interrelated in engagement and retention in the structured lifestyle pro- that participation is always a proportion of penetration grams; and maximizing adherence to the long-term be- and effectiveness can only be attributed to those who haviour changes [59]. participated. Effectiveness, in this context, is defined as the number or proportion of participating cases who Limitations reached a priori defined success criterion. In prevention Several studies did not provide relevant numeric data to of T2DM, diabetes risk reduction is one such success allow the calculation of PIPE coefficients. Hence we criterion. In many studies, however, weight loss was also were unable to compare the overall program impact of a main outcome—either in individual cases as a percent- the included studies. Additionally, any successfully trans- age of overall body weight or across a population as an lated diabetes program should ideally be accessible to those average percentage of weight loss. In our review, we most in need and should have some clear relationship with Aziz et al. Implementation Science (2015) 10:172 Page 14 of 17 the health care delivery system, and we were not able to and program implementation strategies for targeting evaluate these two elements in most of the studies included both reach and program participation that also empha- in this review. sizes long-term program adherence. In our analysis of implementation, we only considered To improve the translation of diabetes prevention pro- standard sessions and have not included the extracurric- grams in real-world settings, we suggest a more rigorous ular or optional activities offered by some of the pro- reporting of program elements and components to grams that may have impact on the future adoption of evaluate these programs to assess the practical value these kinds of programs. This was mainly due to the het- [27] of the diabetes prevention programs. In particular, erogeneity in how the extracurricular activities were or- more detailed reporting on the four key PIPE Metric ganized and also how they were reported. However, we components will provide important insights and has the do acknowledge that these ‘spinoff’ activities are likely to potential to facilitate increased uptake of T2DM prevention be very important for sustainability and the wider diffu- programs worldwide. We also suggest a greater consistency sion of programs, their maintenance, and sustainability. of reporting main outcomes and a standardization of Although there has been a lot of emphasis on the need reporting criteria for translational diabetes prevention pro- for translating T2DM programs in low and middle in- grams implemented in real-world settings. come countries where the majority of people in the world at high risk of diabetes and its progression live Conclusions [60], we did not find any studies from such countries Our findings based on program implementation over the that met our eligibility criteria. The Kerala Diabetes Pre- period of 15 years suggest that while a high-frequency vention Program [61] in India is one of the first such im- intervention plays an important role in achieving high plementation trials to evaluate a peer-led, group-based weight loss outcomes, programs with ‘low’-intense inter- lifestyle intervention program (based on the Finnish ventions have also shown high reductions in the inci- GOAL study) [41] among individuals at high risk of de- dence of T2DM. This suggests that even when the veloping T2DM in rural India. The trial is currently be- effectiveness of an intervention is moderate in terms of ing implemented and will provide important reference weight loss, it can have a profound impact on the devel- to the translation of diabetes prevention programs in opment of a disease at the population level—provided India and similar countries. enough effort is put into guaranteeing high penetration Also, we did not find any studies that have previously and participation as well. From a translation perspective, utilized the PIPE Impact Metric framework in diabetes not many studies provide the necessary information to prevention. More research is needed to understand and estimate the overall impact of such programs. Key ele- apply the few but essential elements of the PIPE model ments of the PIPE Impact Metric are not routinely re- to measure the overall public health impact of diabetes ported in many published implementation trials of prevention interventions. diabetes prevention which therefore reduces their utility for information resource allocation and ‘real-world’ im- Implications for practice plementation. More rigorous evaluation methods are re- The original US-DPP [2] and Finnish-DPS [3] efficacy quired to better understand the factors that influence trials demonstrated that lifestyle intervention is an ef- the likely success of such interventions in the future. fective way to reduce the risk of T2DM in high-risk adults. However, achieving better translation of these Additional files programs still remains challenging after 15 years of re- search. Our findings suggest that program planners and Additional file 1: Search Strategy. Search strategy used for identifying implementers should aim to design high-intensity pro- included studies. (PDF 371 KB) gram with frequent contacts if the primary target is Additional file 2: Table S4–Table S7 Detailed scoring of all weight loss. However, if the primary aim is diabetes risk included studies based on the elements of the PIPE Impact Metric reduction, this can also be achieved with lower fre- framework. (PDF 772 KB) quency of contacts but with a program duration of at least 12 months. With this program design, program Competing interests planners should expect only low or moderate weight The authors declare that they have no competing interests. loss. To have a broader public health impact, programs with lower frequency of contacts but with a program Authors’ contributions duration of at least 12 months might be more feasible The ratings of the included studies were independently derived by three authors (ZA, PA, JO). The ratings of all studies were then reviewed by all but this requires program strategies that simultaneously authors and when disagreement on rating occurred they were resolved with address both penetration and participation. Future trans- consensus among all authors. ZA drafted the manuscript. All authors read, lational research needs to identify effective recruitment provided feedback, and approved the final manuscript. Aziz et al. Implementation Science (2015) 10:172 Page 15 of 17 Authors’ information subjects with impaired glucose tolerance (IDPP-1). Diabetologia. Zahra Aziz MPH is a PhD candidate in the School of Population and Global 2006;49(2):289–97. Health University of Melbourne in Australia. Her PhD is on an evaluation of 6. Diabetes Prevention Program Research Group. 10-year follow-up of diabetes the translation of diabetes prevention programs in ‘real-world’ settings with incidence and weight loss in the diabetes prevention program outcomes a particular focus on the National Health and Medical Research Council study. The Lancet. 2009;374(9702):1677–86. (NHMRC) funded study titled the Kerala Diabetes Prevention program based 7. Li G, Zhang P, Wang J, Gregg EW, Yang W, Gong Q, et al. The long-term effect in rural India. She is also involved in evaluating the Australasian Peers for of lifestyle interventions to prevent diabetes in the China Da Qing diabetes Progress trial in Victoria, Australia. Before commencing her PhD, she has prevention study: a 20-year follow-up study. Lancet. 2008;371(9626):1783–9. worked as an evaluation specialist with the Aga Khan Health Service, 8. Peters DH, Adam T, Alonge O, Agyepong IA, Tran N. Implementation Pakistan for 5 years and has also completed a Master of Public Health research: what it is and how to do it. Bmj. 2013;347:f6753. degree from the University of Melbourne, Australia. She has led the 9. Jensen D. In: Given LM, editor. “Transferability.” The SAGE encyclopedia of conducting of this systematic review. qualitative research methods. Thousand Oaks: SAGE Publications, Inc; Pilvikki Absetz PhD is an Adjunct Professor of Health Promotion at University 2008. p. 887. SAGE knowledge. Web. 12 Mar. 2015. of Tampere, School of Health Sciences, Finland. She is a global expert in lifestyle 10. Yates T, Davies M, Khunti K. Preventing type 2 diabetes: can we make the interventions in prevention and management of type 2 diabetes and other evidence work? Postgraduate medical journal. 2009;85(1007):475–80. chronic conditions. She has been a lead or co-investigator of the following 11. Yamaoka K, Tango T. Efficacy of lifestyle education to prevent type 2 major studies that are directly relevant to type 2 diabetes: the Finnish GOAL diabetes: a meta-analysis of randomized controlled trials. Diabetes care. Lifestyle Implementation Trial; the TERVA tele health coaching intervention trial 2005;28(11):2780–6. in Finland; the Greater Green Triangle DPP in Australia; the Peers for Progress 12. Gillies CL, Abrams KR, Lambert PC, Cooper NJ, Sutton AJ, Hsu RT, et al. trial for diabetes self-management in Australia; and the Kerala DPP in India. She Pharmacological and lifestyle interventions to prevent or delay type 2 is also involved in the European Community Horizon 2020 funded SMART2D diabetes in people with impaired glucose tolerance: systematic review and project promoting cross learning in diabetes prevention and management in meta-analysis. Bmj. 2007;334(7588):299. Sweden, South Africa, and Uganda. 13. Orozco LJ, Buchleitner AM, Gimenez-Perez G, Roque IFM, Richter B, Mauricio John Oldroyd PhD is a research fellow in the Department of Epidemiology D. Exercise or exercise and diet for preventing type 2 diabetes mellitus. The and Preventive Medicine, Monash University. His PhD was in lifestyle Cochrane database of systematic reviews. 2008;3:CD003054. interventions for the prevention of diabetes in people with impaired glucose 14. Angermayr L, Melchart D, Linde K. Multifactorial lifestyle interventions in the tolerance in the UK. primary and secondary prevention of cardiovascular disease and type 2 Nicolaas P. Pronk PhD is vice president and chief science officer at diabetes mellitus—a systematic review of randomized controlled trials. HealthPartners and a researcher in the HealthPartners Institute for Education Annals of behavioral medicine: a publication of the Society of Behavioral and Research in Minneapolis, Minnesota, USA. He is also an adjunct professor Medicine. 2010;40(1):49–64. of Social and Behavioral Sciences at the Harvard School of Public Health, 15. Baker MK, Simpson K, Lloyd B, Bauman AE, Singh MA. Behavioral strategies Boston, MA. He undertakes research in the area of health behaviour change in diabetes prevention programs: a systematic review of randomized and population health. He developed the PIPE model that is used in our controlled trials. Diabetes research and clinical practice. 2011;91(1):1–12. systematic review and is an expert in the evaluation of complex 16. Gillett M, Royle P, Snaith A, Scotland G, Poobalan A, Imamura M, et al. Non- interventions. pharmacological interventions to reduce the risk of diabetes in people with Brian Oldenburg PhD is Professor and Chair of Non-Communicable Disease impaired glucose regulation: a systematic review and economic evaluation. Control in the School of Population and Global Health, University of Mel- Health technology assessment. 2012;16(33):1–236. iii-iv. bourne, Australia. He is a behavioural scientist and an expert in the preven- 17. Schellenberg ES, Dryden DM, Vandermeer B, Ha C, Korownyk C. Lifestyle tion and control of chronic conditions. His research program focuses on interventions for patients with and at risk for type 2 diabetes: a systematic improving the prevention and control of cardiometabolic conditions such as review and meta-analysis. Annals of internal medicine. 2013;159(8):543–51. diabetes, heart disease, and their co-morbidities in Australia and other 18. Terranova CO, Brakenridge CL, Lawler SP, Eakin EG, Reeves MM. countries. Effectiveness of lifestyle-based weight loss interventions for adults with type 2 diabetes: a systematic review and meta-analysis. Diabetes Obes Metab. Author details 2015;17(4):371–8. doi: 10.1111/dom.12430. Epub 2015 Jan 14. Melbourne School of Population and Global Health, University of 19. Stevens JW, Khunti K, Harvey R, Johnson M, Preston L, Woods HB, et al. Melbourne, Melbourne 3010, Australia. School of Health Sciences, University Preventing the progression to type 2 diabetes mellitus in adults at high risk: of Tampere, Tampere FI-33014, Finland. Collaborative Care Systems Finland, a systematic review and network meta-analysis of lifestyle, pharmacological Helsinki, Finland. Department of Epidemiology and Preventive Medicine, and surgical interventions. Diabetes Res Clin Pract. 2015;107(3):320–31. doi: Monash University, Melbourne 3004, Australia. HealthPartners Institute for 10.1016/j.diabres.2015.01.027. Epub 2015 Jan 21. Education and Research, 8170 33rd Ave. S, Minneapolis, MN 55425, USA. 20. Yuen A, Sugeng Y, Weiland TJ, Jelinek GA. Lifestyle and medication interventions for the prevention or delay of type 2 diabetes mellitus in Received: 29 April 2015 Accepted: 20 November 2015 prediabetes: a systematic review of randomised controlled trials. Australian and New Zealand journal of public health. 2010;34(2):172–8. 21. Cardona-Morrell M, Rychetnik L, Morrell SL, Espinel PT, Bauman A. Reduction of diabetes risk in routine clinical practice: are physical activity and nutrition References interventions feasible and are the outcomes from reference trials replicable? 1. International Diabetes Federation. IDF diabetes atlas. 6th ed. Brussels: A systematic review and meta-analysis. BMC public health. 2010;10:653. International Diabetes Federation; 2013. http://www.idf.org/diabetesatlas. 22. Taylor J, Cottrell C, Chatterton H, Hill J, Hughes R, Wohlgemuth C, et al. 2. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker Identifying risk and preventing progression to type 2 diabetes in vulnerable EA, et al. Diabetes Prevention Program Research Group. Reduction in the and disadvantaged adults: a pragmatic review. Diabetic medicine : a journal incidence of type 2 diabetes with lifestyle intervention or metformin. The of the British Diabetic Association. 2013;30(1):16–25. New England journal of medicine. 2002;346(6):393–403. 23. Li R, Zhang P, Barker LE, Chowdhury FM, Zhang X. Cost-effectiveness of 3. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, Ilanne- interventions to prevent and control diabetes mellitus: a systematic review. Parikka P, et al. Prevention of type 2 diabetes mellitus by changes in lifestyle Diabetes care. 2010;33(8):1872–94. among subjects with impaired glucose tolerance. The New England journal 24. Jackson L. Translating the diabetes prevention program into practice: a review of medicine. 2001;344(18):1343–50. of community interventions. The Diabetes educator. 2009;35(2):309–20. 4. Pan XR, Li GW, Hu YH, Wang JX, Yang WY, An ZX, et al. Effects of diet and 25. Ali MK, Echouffo-Tcheugui J, Williamson DF. How effective were lifestyle exercise in preventing NIDDM in people with impaired glucose tolerance. interventions in real-world settings that were modeled on the diabetes The Da Qing IGT and diabetes study. Diabetes care. 1997;20(4):537–44. prevention program? Health affairs. 2012;31(1):67–75. 5. Ramachandran A, Snehalatha C, Mary S, Mukesh B, Bhaskar AD, Vijay V, et al. The Indian diabetes prevention programme shows that lifestyle 26. Johnson M, Jones R, Freeman C, Woods HB, Gillett M, Goyder E, et al. Can modification and metformin prevent type 2 diabetes in Asian Indian diabetes prevention programmes be translated effectively into real-world Aziz et al. Implementation Science (2015) 10:172 Page 16 of 17 settings and still deliver improved outcomes? A synthesis of evidence. Diabetic 47. Sepah SC, Jiang L, Peters AL. Translating the diabetes prevention program medicine : a journal of the British Diabetic Association. 2013;30(1):3–15. into an online social network: validation against CDC standards. The 27. Dunkley AJ, Bodicoat DH, Greaves CJ, Russell C, Yates T, Davies MJ, et al. Diabetes educator. 2014;40(4):435–43. Diabetes prevention in the real world: effectiveness of pragmatic lifestyle 48. Kumanyika S, Fassbender J, Phipps E, Tan-Torres S, Localio R, Morales KH, interventions for the prevention of type 2 diabetes and of the impact of et al. Design, recruitment and start up of a primary care weight loss trial adherence to guideline recommendations: a systematic review and meta- targeting African American and Hispanic adults. Contemporary Clinical analysis. Diabetes care. 2014;37(4):922–33. Trials. 2011;32(2):215–24. 28. Pronk NP, Remington PL; Community Preventive Services Task Force. 49. Duijzer G, Haveman-Nies A, Jansen SC, ter Beek J, Hiddink GJ, Feskens EJ. Combined diet and physical activity promotion programs for prevention of Feasibility and potential impact of the adapted SLIM diabetes prevention diabetes: community preventive services task force recommendation intervention in a Dutch real-life setting: the SLIMMER pilot study. Patient statement. Ann Intern Med. 2015; 163. doi:10.7326/M15-1029. education and counseling. 2014;97(1):101–7. 29. Whittemore R. A systematic review of the translational research on the diabetes 50. Kanaya AM, Santoyo-Olsson J, Gregorich S, Grossman M, Moore T, Stewart prevention program. Translational behavioral medicine. 2011;1(3):480–91. AL. The live well, be well study: a community-based, translational lifestyle program to lower diabetes risk factors in ethnic minority and lower- 30. Glasgow RE. Evaluation of theory-based interventions: the RE-AIM model. In: socioeconomic status adults. Am J Public Health. 2012;102(8):1551–8. Glanz K, Lewis FM, Rimer BK, editors. Health behavior and health education. 3rd ed. San Francisco: Jossey-Bass; 2002. p. 531–44. 51. Paulweber B, Valensi P, Lindstrom J, et al. A European evidence-based 31. Laws RA et al. A systematic review of external validity in lifestyle guideline for the prevention of type 2 diabetes. Horm Metab Res. interventions. Am J Prev Med. 2012;43(2):205–14. 2010;42 Suppl 1:S3–36. 32. Green LW, Glasgow RE. Evaluating the relevance, generalization, and 52. National Institute for Health and Clinical Excellence. Preventing type applicability of research—issues in external validation and translation 2 diabetes: risk identification and interventions for individuals at methodology. Eval Health Prof. 2006;29(1):126–53. high risk. London: National Institute for Health and Clinical Excellence; 2012. 33. Pronk NP. Designing and evaluating health promotion programs: simple 53. Balk EM, Earley A, Raman G, Avendano EA, Pittas AG, Remington PL, Task rules for a complex issue. Disease Management & Health Outcomes. Force on Community Preventive Services. Combined diet and physical 2003;11(3):149–57. activity promotion programs to prevent type 2 diabetes among persons at 34. Zyriax BC, Letsch B, Stock S, Windler E. DELIGHT (delay of impaired glucose increased risk: a systematic review for the community preventive services tolerance by a healthy lifestyle trial)—a feasibility study on implementing a task force. Ann Intern Med. 2015;163(6):437–51 doi: 10.7326/M15-0452. program of sustainable diabetes prevention in German companies. Experimental and clinical endocrinology & diabetes: official journal, 54. Li R, Qu S, Zhang P, Chattopadhyay S, Gregg EW, Albright A, Hopkins D, German Society of Endocrinology [and] German Diabetes Association. Pronk NP, [Task Force on Community Preventive Services]. Economic 2014;122(1):20–6. evaluation of combined diet and physical activity promotion programs to 35. Mensink M, Feskens EJ, Saris WH, De Bruin TW, Blaak EE. Study on lifestyle prevent type 2 diabetes among persons at increased risk: a systematic intervention and impaired glucose tolerance Maastricht (SLIM): preliminary review for the community preventive services task force. Ann Intern Med. results after one year. International journal of obesity and related metabolic 2015; 163. doi:10.7326/M15-0469. disorders: journal of the International Association for the Study of Obesity. 55. Lindstrom J, Peltonen M, Eriksson JG, Louheranta A, Fogelholm M, Uusitupa 2003;27(3):377–84. M, et al. High-fibre, low-fat diet predicts long-term weight loss and decreased type 2 diabetes risk: the Finnish diabetes prevention study. 36. Bo S, Ciccone G, Baldi C, Benini L, Dusio F, Forastiere G, et al. Effectiveness Diabetologia. 2006;49:912–20. of a lifestyle intervention on metabolic syndrome. A randomized controlled 56. Yajnik CS, Lubree HG, Rege SS, Naik SS, Deshpande JA, Deshpande SS, et al. trial. J Gen Intern Med. 2007;22(12):1695–703. Adiposity and hyperinsulinemia in Indians are present at birth. The Journal 37. Lakerveld J, Bot S, Chinapaw M, van Tulder M, Kingo L, Nijpels G. Process of clinical endocrinology and metabolism. 2002;87(12):5575–80. evaluation of a lifestyle intervention to prevent diabetes and cardiovascular diseases in primary care. Health promotion practice. 2012;13(5):696–706. 57. Chan JC, Malik V, Jia W, Kadowaki T, Yajnik CS, Yoon KH, et al. Diabetes 38. Ma J, Yank V, Xiao L, Lavori PW, Wilson SR, Rosas LG, et al. Translating the in Asia: epidemiology, risk factors, and pathophysiology. Jama. diabetes prevention program lifestyle intervention for weight loss into primary 2009;301(20):2129–40. care: a randomized trial. JAMA internal medicine. 2013;173(2):113–21. 58. Resnicow K, McMaster F. Motivational interviewing: moving from why to 39. Davis-Smith YM, Boltri JM, Seale JP, Shellenberger S, Blalock T, Tobin B. how with autonomy support. The international journal of behavioral Implementing a diabetes prevention program in a rural African-American nutrition and physical activity. 2012;9:19. church. Journal of the National Medical Association. 2007;99(4):440–6. 59. Spring B, Sohn MW, Locatelli SM, Hadi S, Kahwati L, Weaver FM. Individual, facility, and program factors affecting retention in a national weight 40. Costa B, Barrio F, Cabre JJ, Pinol JL, Cos X, Sole C, et al. Delaying management program. BMC public health. 2014;14:363. progression to type 2 diabetes among high-risk Spanish individuals is 60. Daivadanam M, Absetz P, Sathish T, Thankappan KR, Fisher EB, Philip feasible in real-life primary healthcare settings using intensive lifestyle NE, et al. Lifestyle change in Kerala, India: needs assessment and intervention. Diabetologia. 2012;55(5):1319–28. planning for a community-based diabetes prevention trial. Bmc Public 41. Absetz P, Valve R, Oldenburg B, Heinonen H, Nissinen A, Fogelholm M, et al. Health. 2013;13:95. Type 2 diabetes prevention in the “real world”: one-year results of the GOAL implementation trial. Diabetes Care. 2007;30(10):2465–70. 61. Sathish T, Williams ED, Pasricha N, Absetz P, Lorgelly P, Wolfe R, et al. Cluster 42. Nilsen V, Bakke PS, Gallefoss F. Effects of lifestyle intervention in persons at randomised controlled trial of a peer-led lifestyle intervention program: risk for type 2 diabetes mellitus—results from a randomised, controlled trial. study protocol for the Kerala diabetes prevention program. BMC public BMC public health. 2011;11:893. health. 2013;13:1035. 62. Kosaka K, Noda M, Kuzuya T. Prevention of type 2 diabetes by lifestyle 43. Penn L, Lordon J, Lowry R, Smith W, Mathers JC, Walker M, et al. Translating intervention: a Japanese trial in IGT males. Diabetes research and clinical research evidence to service provision for prevention of type 2 diabetes: practice. 2005;67(2):152–62. development and early outcomes of the ‘new life, new you’ intervention. 63. Oldroyd JC, Unwin NC, White M, Mathers JC, Alberti KG. Randomised British Journal of Diabetes and Vascular Disease. 2011;11(4):175–81. controlled trial evaluating lifestyle interventions in people with impaired 44. Savas LA, Grady K, Cotterill S, Summers L, Boaden R, Gibson JM. Prioritising glucose tolerance. Diabetes Res Clin Pract. 2006;72(2):117–27. prevention: implementation of IGT care call, a telephone based service for people at risk of developing type 2 diabetes. Primary care diabetes. 2015;9(1):3–8. 64. Laatikainen T, Dunbar JA, Chapman A, Kilkkinen A, Vartiainen E, 45. Vanderwood KK, Hall TO, Harwell TS, Butcher MK, Helgerson SD, Montana Heistaro S, et al. Prevention of type 2 diabetes by lifestyle Cardiovascular D, et al. Implementing a state-based cardiovascular disease intervention in an Australian primary health care setting: greater and diabetes prevention program. Diabetes Care. 2010;33(12):2543–5. green triangle (GGT) diabetes prevention project. Bmc Public Health. 2007;7:249. 46. Janus ED, Best JD, Davis-Lameloise N, Philpot B, Hernan A, Bennett CM, 65. Ackermann RT, Finch EA, Brizendine E, Zhou H, Marrero DG. Translating the et al. Scaling-up from an implementation trial to state-wide coverage: diabetes prevention program into the community. The DEPLOY pilot study. results from the preliminary Melbourne diabetes prevention study. Trials. American Journal of Preventive Medicine. 2008;35(4):357–63. 2012;13:152. Aziz et al. Implementation Science (2015) 10:172 Page 17 of 17 66. Boltri JM, Davis-Smith YM, Seale JP, Shellenberger S, Okosun IS, Cornelius ME. Diabetes prevention in a faith-based setting: results of translational research. Journal of Public Health Management and Practice. 2008;14(1):29–32. 67. Payne WR, Walsh KJ, Harvey JT, Livy MF, McKenzie KJ, Donaldson A, et al. Effect of a low-resource-intensive lifestyle modification program incorporating gymnasium-based and home-based resistance training on type 2 diabetes risk in Australian adults. Diabetes Care. 2008;31(12):2244–50. 68. Kramer MK, Kriska AM, Venditti EM, Miller RG, Brooks MM, Burke LE, et al. Translating the diabetes prevention program. A comprehensive model for prevention training and program delivery. American Journal of Preventive Medicine. 2009;37(6):505–11. 69. Kulzer B, Hermanns N, Gorges D, Schwarz P, Haak T. Prevention of diabetes self-management program (PREDIAS): effects on weight, metabolic risk factors, and behavioral outcomes. Diabetes Care. 2009;32(7):1143–6. 70. Penn L, White M, Oldroyd J, Walker M, Alberti KG, Mathers JC. Prevention of type 2 diabetes in adults with impaired glucose tolerance: the European Diabetes Prevention RCT in Newcastle upon Tyne, UK. BMC Public Health. 2009;9:342. 71. Almeida FA, Shetterly S, Smith-Ray RL, Estabrooks PA. Reach and effectiveness of a weight loss intervention in patients with prediabetes in Colorado. Preventing Chronic Disease. 2010;7(5):A103. 72. Makrilakis K, Liatis S, Grammatikou S, Perrea D, Katsilambros N. Implementation and effectiveness of the first community lifestyle intervention programme to prevent type 2 diabetes in Greece. The DE- PLAN study. Diabetic Medicine. 2010;27(4):459–65. 73. Parikh P, Simon EP, Fei K, Looker H, Goytia C, Horowitz CR. Results of a pilot diabetes prevention intervention in East Harlem, New York City: project HEED. American journal of public health. 2010;100 Suppl 1:S232–9. 74. Vermunt PW, Milder IE, Wielaard F, van Oers JA, Westert GP. An active strategy to identify individuals eligible for type 2 diabetes prevention by lifestyle intervention in Dutch primary care: the APHRODITE study. Family practice. 2010;27(3):312–9. 75. Boltri JM, Davis-Smith M, Okosun IS, Seale JP, Foster B. Translation of the National Institutes of Health diabetes prevention program in African American churches. Journal of the National Medical Association. 2011;103(3):194–202. 76. Gilis-Januszewska A, Szybinski Z, Kissimova-Skarbek K, Piwonska-Solska B, Pach D, Topor-Madry R, et al. Prevention of type 2 diabetes by lifestyle intervention in primary health care setting in Poland: diabetes in Europe prevention using lifestyle, physical activity and nutritional intervention (DE-PLAN) project. Br J Diabetes Vasc Dis. 2011;11(4). 77. Katula JA, Vitolins MZ, Rosenberger EL, Blackwell CS, Morgan TM, Lawlor MS, et al. One-year results of a community-based translation of the diabetes prevention program: healthy-living partnerships to prevent diabetes (HELP PD) project. Diabetes Care. 2011;34(7):1451–7. 78. Ruggiero L, Oros S, Choi YK. Community-based translation of the diabetes prevention program's lifestyle intervention in an underserved Latino population. Diabetes Educator. 2011;37(4):564–72. 79. Sakane N, Sato J, Tsushita K, Tsujii S, Kotani K, Tsuzaki K, et al. Prevention of type 2 diabetes in a primary healthcare setting: three-year results of lifestyle intervention in Japanese subjects with impaired glucose tolerance. BMC Public Health. 2011;11(1):40. 80. Ockene IS, Tellez TL, Rosal MC, Reed GW, Mordes J, Merriam PA, et al. Outcomes of a Latino community-based intervention for the prevention of diabetes: the Lawrence Latino diabetes prevention project. American Journal of Public Health. 2012;102(2):336–42. 81. Piatt GA, Seidel MC, Chen H-Y, Powell RO, Zgibor JC. Two-year results of translating the diabetes prevention program into an urban, underserved community. Diabetes Educator. 2012;38(6):798–804. Submit your next manuscript to BioMed Central 82. Jiang LH, Manson SM, Beals J, Henderson WG, Huang HX, Acton KJ, et al. and we will help you at every step: Translating the diabetes prevention program into American Indian and Alaska native communities results from the special diabetes program for • We accept pre-submission inquiries Indians diabetes prevention demonstration project. Diabetes Care. � Our selector tool helps you to find the most relevant journal 2013;36(7):2027–34. � We provide round the clock customer support � Convenient online submission � Thorough peer review � Inclusion in PubMed and all major indexing services � Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit

Journal

Implementation Science – Springer Journals

Published: Dec 15, 2015

There are no references for this article.