NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Gudzune KA, Lau BD, Hutfless S, et al. Strategies To Prevent Weight Gain in Adults: Future Research Needs: Identification of Future Research Needs From Comparative Effectiveness Review No. 97 [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2013 Jun. (Future Research Needs Papers, No. 43.)

Cover of Strategies To Prevent Weight Gain in Adults: Future Research Needs

Strategies To Prevent Weight Gain in Adults: Future Research Needs: Identification of Future Research Needs From Comparative Effectiveness Review No. 97 [Internet].

Show details


We identified potential research needs based on research gaps identified while writing the report and prioritized with input from stakeholders. The protocol for developing the evidence gaps into a prioritized list of research needs and feasible researchable questions involved the following steps: (1) identification of evidence gaps, (2) engagement of stakeholders, (3) prioritization of PICOTS (population, intervention, comparison, outcome, timing, setting) research gaps through the Delphi process, (4) creation and prioritization of research questions through the Delphi process, and (5) identification of ongoing studies through external literature search.

Identification of Evidence Gaps

To identify evidence gaps, our research team abstracted evidence gaps from the draft CER “Strategies To Prevent Weight Gain Among Adults.”10 The report authors identified evidence gaps based on the strength of evidence, applicability, and limitations of the review. We also identified additional evidence gaps and limitations of the review during in-person discussions with the original report authors. We created an intervention matrix of comparisons for each Key Question.

We considered all findings with low or insufficient strength of evidence as evidence gaps. Given that the report authors graded almost all evidence as low or insufficient to address the Key Questions (Table 2), all research questions were identified as gaps in the literature. This extensive deficiency made the use of the typical analytic framework for identifying future research needs difficult.11 A prior report also noted this challenge in a similar situation,12 and we modeled our approach based on its methods. We opted to have our stakeholders prioritize different PICOTS gaps (populations, interventions or comparisons of interventions to each other, outcomes, timing of interventions, settings) for future research needs. We then used this prioritization to create a list of questions for future research.

Table 2. Summary of the strength of evidence from comparative effectiveness review.

Table 2

Summary of the strength of evidence from comparative effectiveness review.

Criteria for Prioritization

We used the Delphi method to prioritize and develop consensus about future research needs.13 After reading the Executive Summary of the draft CER, each stakeholder was asked to select the highest and lowest priority populations, interventions/strategies, comparisons, outcomes, and settings. Stakeholders were asked to respond based on their reading of the executive summary and the potential health impact of each element. We defined consensus as a majority of respondents identifying an element as being among the highest priority.

Delphi Round 1. Prioritization of Populations, Interventions, Outcomes, and Settings

Populations. Stakeholders identified four populations that they felt were the highest priority from a list of all populations in the CER and populations not represented in the report such as age, sex, race/ethnicity, or socioeconomic status. The list of possible populations included: all adults, adults with cardiovascular disease/diabetes, adults with severe mental illness, adults with cancer, young adults (age 18–35), middle-aged adults (age 36–64), older adults (age >64), women, men, low income, racial/ethnic minorities, normal weight, overweight, obese, or other.

Interventions/strategies. Stakeholders selected two interventions/strategies that they felt were the highest priority from a list of all possible interventions proposed in the Key Questions of the CER: self-management, diet, physical activity, medication, environmental/policy, a combination of these interventions, or other. The stakeholders also chose the lowest priority intervention/strategy from this same list.

Outcomes. Given the widespread usage of weight and body mass index to study weight gain prevention, we opted to ask stakeholders to prioritize the top two secondary outcomes in this field of research. We created a list of secondary outcomes based on elements of the analytic framework from the CER and other outcomes noted by the report authors to be commonly assessed throughout the literature. These secondary outcomes included waist circumference, body fat percentage, skinfold thickness, adherence, adverse effects, weight-related clinical conditions, mortality, or other. Stakeholders selected the one lowest priority secondary outcome from this same list.

Settings. Stakeholders identified the two highest priority settings for future research from a list of settings identified from the CER. The list included college, clinic, work, community, home, or other. Stakeholders also selected the lowest priority setting from this same list.

Delphi Round 2. Prioritization of Populations, Interventions, and Comparisons

Populations. In round 1, several population subgroups received few or no votes from the stakeholders; however, the report authors had identified these populations as potential important subgroups for future research given their high risk of obesity. These subgroups included adults with cardiovascular disease/diabetes, adults with severe mental illness, adults with cancer, racial/ethnic minorities, and obese adults. Therefore, we asked the stakeholders to rank these subgroups for future research in adult weight gain prevention from highest to lowest on a scale of 1–5.

Interventions/strategies. In round 1, stakeholders overwhelmingly identified combination interventions as the highest priority and medications as the lowest priority. To clarify which combination of interventions they would recommend, we asked them to select the two highest priority combination strategies from a list. We created this list based on the possible combinations of self-management plus diet, physical activity, and/or environment. We defined self-management as goal setting, self-monitoring, problem solving, relapse prevention, and stimulus control. We included self-management in all combinations, as our study team felt these targeted behaviors to be an essential part of any behavior change intervention. The stakeholders also selected the lowest priority combination strategy from this same list.

Comparisons. In the CER, few head-to-head comparative studies were described, since virtually all studies compared intervention to control. Based on comments from the report authors and stakeholders in round 1, we asked stakeholders to select the five highest priority comparisons of combination interventions. We selected comparisons identified by more than 50 percent of stakeholders as high-priority comparisons.

Delphi Round 3. Prioritization of Components of Interventions and Research Questions

Interventions/strategies. In round 2, the majority of stakeholders identified the comparison of (1) self-management + diet versus self-management + physical activity and (2) self-management + diet versus self-management + diet + physical activity. Concerning these priority combination interventions, the report authors commented on the heterogeneity of elements included within diet and physical activity interventions. Therefore, we asked stakeholders to select the two highest priority combinations of elements of diet and physical activity interventions. Stakeholders could choose to include a diet/exercise plan and/or intervention and/or targeted behavior. For dietary interventions, dietary plans included changes in dietary composition, caloric restriction, both, or none and targeted behaviors included meal planning, calorie tracking, both, or none. For exercise interventions, exercise plans included all combinations of aerobic exercise, resistance training, and/or stretching and targeted behaviors included exercise tracking, pedometer, both, or none. Stakeholders could select from group counseling, individual counseling, telephone/Web-based counseling, education session, or printed materials for the intervention aspect for both diet and exercise. We created a list of elements based on components identified in the CER.

Concurrent with ranking components of dietary and exercise interventions, Delphi round 3 also asked stakeholders to prioritize the top five questions for future research based on results from Delphi rounds 1 and 2. (See research question development section below.)

Delphi Round 4. Ranking of Prioritized Research Questions

Delphi round 4 asked stakeholders to rate the value of the prioritized research questions based on potential value and impact of results. (See research question development section below.)

Engagement of Stakeholders, Researchers, and Funders

The stakeholders' role was to participate in the prioritization of PICOTS elements and subsequent questions for future research.

Identification and Recruitment of Stakeholders

We wanted to identify and recruit stakeholders that represented a variety of interests. Our team first generated a list of stakeholder groups that would potentially have an interest in the prevention of weight gain in adults, which included patients, physicians, exercise and nutrition researchers, funding agencies, and health insurers. We then identified possible stakeholders within each of these groups who could represent their interests. All possible stakeholders were contacted via email. We invited previous stakeholders and reviewers for the CER, of whom two prior Key Informants agreed to participate. We also invited new participants. Table 3 lists our stakeholders. All participating stakeholders provided curriculum vitae and disclosure statements to ensure that all potential conflicts of interest were disclosed. The list of stakeholders and their disclosure statements were approved by AHRQ.

Table 3. Composition of stakeholders group.

Table 3

Composition of stakeholders group.

Orientation of Stakeholders

All stakeholders received a copy of the executive summary from the draft 2012 CER on strategies to prevent adult weight gain10 and a Web link to the complete draft report. We requested that stakeholders read the Executive Summary in order to meaningfully contribute to the Delphi process identifying future research needs. We informed stakeholders that we anticipated four rounds of surveys, which would be administered using a Web-based survey tool (Survey Monkey The surveys included both multiple-choice and drop-down menus, as well as comment boxes where stakeholders could add free-text responses to ensure that their input was heard and provide clarification and reasoning for their selections.

Research Question Development

Based upon results from the first rounds of the Delphi process, we created a matrix of possible research questions based on PICOTS elements that stakeholders identified as highest priority through consensus. We continued to use the Delphi process with our same stakeholders to prioritize questions for future research. In Delphi round 3, we presented each stakeholder with the list of research questions created from our matrix. The stakeholders selected the top five research questions based on their potential health impact. Based on the rankings from round 3, we presented the final seven priority research questions to the stakeholders in round 4. We asked each stakeholder to assign a score from 1 to 5 to each question, where 5 indicates that the question is very likely to provide valuable and impactful results and 1 that it is less likely to provide valuable and impactful results. We opted not to integrate the prioritized components of dietary and exercise interventions into our research questions.

Research Design Considerations

With the report authors, we reviewed the research questions that were prioritized by our stakeholders. We asked the report authors to weigh in on the potential research designs that would be most appropriate to answer these research questions, including their rationale and the factors that influenced their decision for choosing such designs over others.

With our stakeholders, we assessed methodologic gaps, which we defined as limitations in study design and reporting elements found within the literature base. We identified these methodologic gaps based upon study design limitations cited in the CER and through in-person discussions with the report authors. During round 1 of the Delphi process, stakeholders were given a list of methodologic issues and indicated which issues were important to incorporate into the design of future research. We created the list of study design and reporting elements from limitations abstracted from the CER and from in-person discussion with the report authors. Stakeholders were asked to base their responses upon the Executive Summary and to take into account factors such as feasibility, time, costs, and validity of incorporating these study design and reporting elements into future research. In addition, we had stakeholders consider whether a standard definition of weight maintenance in adults would benefit the field. Based upon their responses to this question, we created a methodologic question for future research.

Identification of Ongoing Studies

We identified ongoing clinical trials that may address the Key Questions in the Comparative Effectiveness Review by searching the National Institutes of Health's registry, (, for trials registered from January 1, 2008, through October 1, 2012. A single reviewer examined each title for entry. A single reviewer then examined the full information from on these potential trials to determine whether they met the same inclusion/exclusion criteria used in the CER. If any ongoing studies met our inclusion criteria, then a single reviewer abstracted the trial identification number, date of registry, expected date of completion, study name, status, method compared, and any published results, and identified the Key Question that the study is likely to address.

PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (564K)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...