Implementation research on noncommunicable disease prevention and control interventions in low- and middle-income countries: A systematic review.

BACKGROUND: While the evidence for the clinical effectiveness of most noncommunicable disease (NCD) prevention and treatment interventions is well established, care delivery models and means of scaling these up in a variety of resource-constrained health systems are not. The objective of this review was to synthesize evidence on the current state of implementation research on priority NCD prevention and control interventions provided by health systems in low- and middle-income countries (LMICs). METHODS AND
FINDINGS: On January 20, 2021, we searched MEDLINE and EMBASE databases from 1990 through 2020 to identify implementation research studies that focused on the World Health Organization (WHO) priority NCD prevention and control interventions targeting cardiovascular disease, cancer, diabetes, and chronic respiratory disease and provided within health systems in LMICs. Any empirical and peer-reviewed studies that focused on these interventions and reported implementation outcomes were eligible for inclusion. Given the focus on this review and the heterogeneity in aims and methodologies of included studies, risk of bias assessment to understand how effect size may have been compromised by bias is not applicable. We instead commented on the distribution of research designs and discussed about stronger/weaker designs. We synthesized extracted data using descriptive statistics and following the review protocol registered in PROSPERO (CRD42021252969). Of 9,683 potential studies and 7,419 unique records screened for inclusion, 222 eligible studies evaluated 265 priority NCD prevention and control interventions implemented in 62 countries (6% in low-income countries and 90% in middle-income countries). The number of studies published has been increasing over time. Nearly 40% of all the studies were on cervical cancer. With regards to intervention type, screening accounted for 49%, treatment for 39%, while prevention for 12% (with 80% of the latter focusing on prevention of the NCD behavior risk factors). Feasibility (38%) was the most studied implementation outcome followed by adoption (23%); few studies addressed sustainability. The implementation strategies were not specified well enough. Most studies used quantitative methods (86%). The weakest study design, preexperimental, and the strongest study design, experimental, were respectively employed in 25% and 24% of included studies. Approximately 72% of studies reported funding, with international funding being the predominant source. The majority of studies were proof of concept or pilot (88%) and targeted the micro level of health system (79%). Less than 5% of studies report using implementation research framework.
CONCLUSIONS: Despite growth in implementation research on NCDs in LMICs, we found major gaps in the science. Future studies should prioritize implementation at scale, target higher levels health systems (meso and macro levels), and test sustainability of NCD programs. They should employ designs with stronger internal validity, be more conceptually driven, and use mixed methods to understand mechanisms. To maximize impact of the research under limited resources, adding implementation science outcomes to effectiveness research and regional collaborations are promising.

Introduction

Noncommunicable diseases (NCDs) have become the leading contributors to morbidity and mortality worldwide. They are now responsible for 74% of all global deaths, 77% of which occur in low- and middle-income countries (LMICs) [1,2]. Approximately 85% of NCD deaths among people aged 30 and 69 years occur in LMICs [1]. Cardiovascular diseases are the leading causes of NCD mortality, followed by cancers, respiratory diseases, and diabetes [1]. Together, these 4 NCDs are responsible of over 80% of all premature NCD deaths [1]. Risk factors such as tobacco and alcohol use, physical inactivity, and unhealthy diets result in significantly greater risk of dying from NCDs. Primary, secondary, and tertiary prevention strategies are vital in addressing NCD burden [1]. Sustainable Development Goal (SDG) target 3.4 commits countries to reduce premature mortality from NCDs by a third by 2030 relative to 2015 levels. Recent analysis shows that no LMIC is on track to meet this target for both men and women if they maintain their 2010 to 2016 average rates of decline [3].

NCD prevention and control should not be regarded as a vertical issue separated from other health conditions. The ongoing Coronavirus Disease 2019 (COVID-19) pandemic has put a spotlight on NCDs, as these increased the risk of death for people with COVID infection. Similarly, NCDs increase mortality risk among people with other infectious diseases such as tuberculosis and HIV. It further highlighted the economic and social inequities in who is afflicted with NCDs, in both high-income countries and LMICs. While primary prevention relies on public health, taxation, and other public policy measures, mitigating the health consequences of NCDs also requires strong health systems. Health systems that recognize this challenge and address modifiable risk factors and prioritize the management of NCDs will be better positioned to promote and maintain health. Data from the 2019 World Health Organization (WHO) NCD Country Capacity surveys reveal that only half of 160 countries have national guidelines for NCDs, half have the 6 essential technologies for early detection, diagnosis, and monitoring of NCDs available in primary care facilities of the public health sector, and 20% of countries have 6 (or fewer) of the 11 essential medicines available [4]. Greater prioritization of NCDs within health systems and high-quality care are essential to achieving SDG 3.4 [3]. Beyond this lies an important agenda for tackling the cumulatively large group of rarer NCDs that afflict the world’s poorest people [5].

To support countries in crafting effective NCD strategies, the WHO Assembly endorsed the Global Action Plan for the Prevention and Control of Noncommunicable Diseases 2013–2020 (GAP-NCD) in May 2013 together with a set of evidence-based interventions (best-buys) and policy options in its appendix 3 that was updated in 2016 and provides 84 interventions or policy options [6,7]. Furthermore, WHO has developed a compendium including all available health interventions. The list and compendium aim to assist Member States, as appropriate in specific national contexts, in implementing measures to achieve the 9 global voluntary targets for NCDs and Target 3.4 of the SDGs. Despite recent calls for a new commitment to implementation research for NCDs, a mid-point evaluation of the WHO NCD Global Action Plan 2013–2030 (NCD-GAP) found that “research has been the weakest NCD-GAP objective in terms of implementation and that progress in implementing research linked to the NCD-GAP has been slow and incremental” [8,9].

While the evidence for the clinical effectiveness of most NCD prevention and treatment interventions is well established, care delivery models and means of scaling these up to entire populations in need in heterogeneous and resource-constrained health systems are not. Implementation research on NCD program delivery, including cost effectiveness in various regions, can illuminate what does and does not work in achieving NCD control [8,10-12]. This can promote faster, more efficient, and more effective scale-up of life-saving and health-preserving health system strategies [13,14]. In this systematic review, we aim to synthesize evidence on the current state of implementation research on WHO priority NCD prevention and control interventions provided within health systems in LMICs [6,7,15-17].

Methods

This systematic review was conducted according to a study protocol registered in PROSPERO (#CRD42021252969) [18].

Search strategy

Following the Systematic Reviews and Meta-Analyses (PRISMA) checklist [19], we searched for implementation research studies that focused on relevant NCD prevention and control interventions (Table A in S1 Appendix) provided within health systems in LMICs and were published in peer-reviewed journals indexed in MEDLINE and EMBASE databases from 1990 to 2020. The databases were last searched on January 20, 2021. Our search terms included medical subject heading (MeSH) terms and/or key words for 4 key themes (implementation research; NCDs; NCD interventions; LMICs) that were adjusted for each database:

Implementation research (e.g., implementation research, implementation science, diffusion of innovations, implementation strategies, dissemination science, implementation outcomes).

NCDs (e.g., cardiovascular disease, cancer, diabetes, chronic respiratory disease).

Interventions (e.g., smoking cessation, management of hypertension, treatment of acute myocardial infarction, cervical and colorectal cancer screening).

LMICs as defined by the World Bank in 2019 (Table C in S1 Appendix).

Language restrictions were not applied. Full details of the search strategy are provided in Table B in S1 Appendix.

Table 1 summarizes our review’s specific eligibility criteria. This review includes peer-reviewed, empirical quantitative, qualitative, and mixed method study designs conducted in LMICs that described the implementation of relevant NCD preventive and/or control interventions provided within health systems. Using the updated Appendix 3 of the WHO Global NCD Action Plan 2013–2020, we identified the WHO priority NCD prevention and control interventions [6]. Of these interventions, we selected those that are specifically provided by health systems. This was achieved through discussions and consensus. Table 2 summarizes the intervention categories across eligible NCD risk factors (i.e., tobacco and alcohol use, physical inactivity, and unhealthy diets) and NCDs (i.e., cardiovascular disease, diabetes, cancer, and chronic respiratory disease), and full details are provided in Table A in S1 Appendix. While our search in databases was not restricted to any language, during study screening/review processes, we only retained eligible studies that were in 6 official languages of the United Nations (i.e., Arabic, Chinese, English, French, Russian, and Spanish). We drew on Proctor and colleagues and Glasgow and colleagues to define implementation outcomes eligible for inclusion [20,21]. Nonempirical/primary research studies are not eligible for inclusion (Table 1).

Table 1

Inclusion and exclusion criteria.

	Inclusion criteria	Exclusion criteria
Population	Human beings with or without NCDs. Human beings with or without NCD risk factors.	Subjects are not human beings.
Intervention	NCD prevention and/or control interventions that are provided within health systems (see Table A in S1 Appendix).	Interventions that are not specified in the inclusion criteria.
Outcome	Implementation outcomes as defined by Proctor and colleagues and Glasgow and colleagues [20,21]    • Acceptability    • Adoption    • Appropriateness    • Feasibility    • Fidelity    • Penetration    • Sustainability    • Implementation costs    • Reach    • Implementation    • Maintenance	Outcomes other than those specified in the inclusion criteria.
Study design	Quantitative, qualitative, or mixed method.Quantitative study designs included experimental and observational.    • Experimental designs:      ○ Randomized controlled trial,      ○ Cluster randomized trial,      ○ Randomized step wedge,    • Observational designs:      ○ Quasi-experimental designs:        ▪ Single interrupted time series,        ▪ Controlled interrupted time series,        ▪ Pre-post with comparison group,        ▪ Regression discontinuity,        ▪ Nonrandomized stepped wedge      ○ Preexperimental designs (no control group or no repeated measures):        ▪ Pre-post        ▪ Post-only design      ○ Other observational designs include:        ▪ Cohort studies        ▪ Cross-sectional studies        ▪ Case-control studies	Nonempirical/primary research including:    • Review    • Meta-analysis    • Editorial    • Commentary    • Letter to editor    • Opinion paper    • Newspaper    • Protocols    • Case report    • Epidemiological/descriptive studies (e.g., nonintervention association studies including knowledge, attitude, discrete choice experiment, awareness, willingness, and perception (including perceived barriers) studies) and not in the context of implementation of NCD interventions.    • Instrument/screening or diagnostic test validation studies    • Call to action    • Sharing experience/lessons learned on the field if not resulting from research    • (Descriptive) cost-effectiveness studies based on modeling (and not in the context of implementation of NCD interventions)
Geographic Scope	LMICs (see Table C in S1 Appendix)	Areas other than LMICs
Time frame	1990–2020	Studies published before 1990

LMIC, low- and middle-income country; NCD, noncommunicable disease.

Table 2

Summary of eligible NCD preventive and control interventions.

Conditions	Intervention categories
NCD risk factors
Tobacco use	Individual smoking cessation
	Mass media campaign smoking cessation
Harmful use of alcohol	Alcohol reduction counseling for at risk individuals
	Treatment for alcohol use disorder
Unhealthy diet	Mass media or other behavior change program to reduce salt intake
	Nutrition education in institutions
	Salt reduction public institutions
	Interventions to promote exclusive breastfeeding
Physical inactivity	Community environmental program increase physical activity
	Mass media campaign promote physical activity
	Physical activity counseling
NCDs
Cardiovascular disease	Treatment of hypertension
	Rehabilitation of post-acute CVD event (myocardial infarction, stroke)
	Treatment of high-risk CVD event
	Treatment of acute ischemic stroke
	Treatment of acute myocardial infarction
	Treatment of heart failure
	Antibiotic treatment of streptococcal pharyngitis (rheumatic fever prevention)
	Treatment for secondary prevention of stroke (e.g., anticoagulation for atrial fibrillation, aspirin)
Diabetes	Glycemic control among people with diabetes
	Screening to prevent complications among people with diabetes
	Treatment of diabetes
	Preconception care for women with diabetes
	Influenza vaccination for people with diabetes
Cancer	Breast cancer screening
	Cervical cancer screening
	HPV vaccination for teen girls
	Colorectal cancer screening
	Treatment of breast and colorectal cancer
	Hepatitis B immunization for liver cancer prevention
	Screening for oral cancer in high-risk groups
Chronic respiratory disease	Treatment of asthma and COPD
	Influenza vaccination for patients with COPD

COPD, chronic obstructive pulmonary disease; CVD, cardiovascular disease; HPV, human papilloma virus; NCD, noncommunicable disease.

Data extraction and analysis

The titles and abstracts of unique results from the databases were reviewed independently by 2 researchers for potential inclusion using COVIDENCE review software [22]. The full texts of studies retained at the title and abstract screening stage were retrieved and independently assessed for inclusion. Any discrepancies were resolved through discussion and consensus. Data extraction on each included study was conducted by a single researcher using a data extraction tool, developed and piloted a priori (Table D in S1 Appendix). Data elements included study characteristics (e.g., publication year, country of implementation, study funding), NCD conditions (risk factors and disease), intervention details (e.g., type of intervention, level of health system), methods (e.g., research approach, study design), implementation outcomes (e.g., fidelity, feasibility), and equity lens (e.g., disaggregated by key SES stratifiers, targeted vulnerable population). We also extracted data on implementation strategies including actor (i.e., who delivered the intervention), action target, and recipients; details of other implementation strategies were not sufficiently described to permit extraction [23]. The recipients of the action/strategy were further aggregated by demographic subgroup (e.g., people eligible for cancer screening including cervical and colorectal), disease risk subgroup (e.g., patients with myocardial infarction, patients with diabetes or hypertension, people who smoke), general population, healthcare workers (e.g., physicians, nurses, pharmacists, and midwives), and community health workers. We synthesize extracted data using descriptive statistics and following the review protocol registered in PROSPERO. Specifically, we provide an overview of NCD priority intervention implementation study characteristics across NCD conditions to shed light on the current state of implementation research of priority NCD prevention and control interventions in LMICs. Given this review does not focus on effect size of NCD interventions, we did not perform a meta-analysis.

Risk of bias assessment

This review focuses on implementation of multiple interventions across various NCDs, rather than effectiveness of any single set of interventions. Further, studies with heterogenous aims and methodologies (including qualitative methodology) were included. Therefore, risk of bias assessment to understand how effect size may have been compromised by bias is not applicable in this review. We instead commented on the distribution of research designs and discussed about stronger/weaker designs.

Results

Our search strategy implemented in MEDLINE and EMBASE identified 9,683 publications, of which 7,419 unique records were screened for inclusion. Abstract and full-text screening identified 222 studies that met our inclusion criteria (Tables 1 and 2) [24-245]. A summary of this process is presented in the PRISMA flow diagram in Fig 1.

Fig 1

PRISMA flow chart.

Intervention refers to studies excluded because they studied the implementation of interventions that did not meet the eligible criteria. Study design refers to studies excluded because they used study designs that did not meet eligibility criteria (e.g., nonempirical studies including reviews and commentaries). Outcomes refer to studies excluded based on not having focused on relevant implementation outcomes. Settings refer to studies excluded because they were not conducted in LMICs. Full text means that studies were excluded because full text was not available. Time refers to studies that were excluded because they were published before/conducted before 1990.

The 222 studies included in this review evaluated 265 priority NCD prevention and control interventions implemented in 62 countries, of which 6% were in low-income countries (LICs), 45% in LMICs, and 46% in upper middle-income countries (UMICs) (Table 3 and Figs 2, 3, and 4A and Table E in S1 Appendix). The NCD conditions targeted varied by income groups of countries (Fig A in S1 Appendix). Eight of the included studies were multicountry studies. The number of studies published has been increasing over time (Fig 5A). Overall, the majority of interventions were focused on either screening (49%) or treatment (39%), while prevention accounted for only 12%, with nearly 80% of these tackling prevention of the shared NCD behavioral risk factors—tobacco use, unhealthy diet, physical inactivity, and harmful use of alcohol. The NCD interventions varied by conditions and type (prevention, screening, and treatment) (Figs 2, B, and C in S1 Appendix). Notably, over one-third of the interventions studied (37%) were for cervical cancer (Fig 2), which accounts for 0.35% of DALYs lost and 0.5% of deaths globally, with similar figures for LMICs (https://vizhub.healthdata.org/gbd-compare/). Diabetes was the focus of nearly one-quarter of the research with hypertension the topic of another 9% (Fig 2). Each of the other recommended interventions represented 5% or less of the implementation research output. Chronic respiratory disease was understudied relative to its prevalence: less than 1% of the studies examined chronic respiratory disease treatment and only 3% smoking cessation programs. The intervention focus appears to vary by income groups of countries (Fig D in S1 Appendix). Feasibility was the most studied implementation outcome followed by adoption (Fig 6). Most of the actors were researchers, which accounted for 58%; whereas government/ministry of health, providers, and NGOs accounted for 18%, 10%, and 6%, respectively. The majority of intervention targeted improvement in health outcomes (45%) followed by change in behavior (34%).

Table 3

Overview of study characteristics.

NCDs and risk factors	Intervention categories	N	Distribution of priority NCD interventions (N = 265)
			Country’s income classification, N	Methods approach, N	Major study design, N	Health system level*, N	Level of scale-up, N	Implementation outcomes, N	Considered equity†, N	Implementation strategies
										Actor, N	Action target, N	Recipients, N
Tobacco use	Individual smoking cessation	6	LMICs = 5UMICs = 1	Quantitative = 5Mixed = 1	Experimental = 2Multiple = 1Preexperimental = 2Other Observational = 1	Micro = 4Meso = 2	Pilot† = 5Scale-up = 1	Adoption = 1Appropriateness = 1Feasibility = 3Multiple = 1	4	Researchers = 4Providers = 2	Behavior = 6	Disease risk subgroup = 6
	Mass media campaign smoking cessation	2	UMICs = 2	Quantitative = 2	Experimental = 1Observational = 1	Macro = 2	Scale-up = 2	Adoption = 1Penetration = 1	1	Researchers = 1MOH = 1	Behavior = 2	Disease risk subgroup = 2
Harmful use of alcohol	Alcohol reduction	1	LMICs = 1	Quantitative = 1	Experimental = 1	Micro = 1	Pilot = 1	Multiple = 1	0	Researchers = 1	Behavior = 1	Disease risk subgroup = 1
Unhealthy diet	Mass media or other behavior change program to reduce salt intake	3	LMICs = 2UMICs = 1	Quantitative = 3	Experimental = 2Other observational = 1	Micro = 1Meso = 1Macro = 1	Pilot = 1Scale-up = 2	Adoption = 1Penetration = 1Multiple = 1	2	Researchers = 1MOH = 2	Behavior = 3	General population = 2Disease risk subgroup = 1
	Nutrition education in institutions	5	LMICs = 1UMICs = 3Multiple = 1	Quantitative = 4Mixed method = 1	Quasi-experimental designs = 3Preexperimental = 1Other observational = 1	Micro = 1Meso = 3Macro = 1	Pilot = 2Scale-up = 3	Acceptability = 1Adoption = 2Feasibility = 1Penetration = 1	3	Researchers = 4MOH = 1	Behavior = 3Behavior, health outcomes = 2	Demographic subgroup = 2Disease risk subgroup = 3
	Salt reduction public institutions	2	UMICs = 2	Quantitative = 2	Other observational = 2	Macro = 2	Pilot = 1Scale-up = 1	Adoption = 1Penetration = 1	1	Researchers = 1MOH = 1	Behavior = 2	Demographic subgroup = 2
Physical inactivity	Community environmental program increase physical activity	4	LMICs = 3UMICs = 1	Quantitative = 4	Experimental = 2Preexperimental = 1Other observational = 1	Micro = 1Meso = 1Macro = 2	Pilot = 1Scale-up = 3	Feasibility = 1Penetration = 1Multiple = 2	2	Researchers = 3MOH = 1	Behavior = 2Behavior and knowledge = 2	Demographic subgroup = 3Disease risk subgroup = 1
	Mass media campaign promote physical activity	2	UMCIs = 2	Quantitative = 2	Experimental = 1Other observational = 1	Macro = 2	Scale-up = 2	Adoption = 1Penetration = 1	1	Researchers = 1MOH = 1	Behavior = 1Behavior and knowledge = 1	Demographic subgroup = 1Disease risk subgroup = 1
CVD	Rehabilitation post-acute CVD event	1	UMICs = 1	Quantitative = 1	Experimental = 1	Micro = 1	Pilot = 1	Feasibility = 1	0	Researchers = 1	Health outcomes = 1	Disease risk subgroup = 1
	Treatment of high-risk CVD event	5	LMICs = 2UMICs = 3	Quantitative = 5	Experimental = 2Quasi-experimental designs = 2Other observational = 1	Micro = 5	Pilot = 5	Acceptability = 1Adoption = 2Feasibility = 1Maintenance = 1	3	Researchers = 4Providers = 1	Behavior = 3Health outcomes = 1	Demographic subgroup = 1Disease risk subgroup = 3HCWs = 1
	Treatment of acute ischemic stroke	10	LMICs = 5UMICs = 5	Quantitative = 10	Experimental = 2Preexperimental = 7Other observational = 1	Micro = 6Meso = 4	Pilot = 8Scale-up = 2	Adoption = 4Feasibility = 4Fidelity = 1	1	Researchers = 5MOH = 4Providers = 1	Health outcomes = 10	Disease risk subgroup = 10
	Treatment of acute myocardial infarction	12	LMICs = 2UMICs = 10	Quantitative = 11Qualitative = 1	Experimental = 3Quasi-experimental designs = 2Preexperimental = 4Other observational = 3	Micro = 6Macro = 6	Pilot = 10 Scale-up = 1	Adoption = 5Feasibility = 5Fidelity = 1Penetration = 1	3	Researchers = 4MOH = 4Providers = 4	Health outcomes = 11Behavior = 1	Disease risk subgroup = 11HCWs = 1
	Treatment of heart failure	5	LMICs = 2UMICs = 3	Quantitative = 5	Experimental = 2Quasi-experimental designs = 1Preexperimental = 2	Micro = 5	Pilot = 5	Adoption = 2Feasibility = 3	1	Researchers = 3Providers = 2	Health outcomes = 5	Disease risk subgroup = 5
	Treatment of hypertension	23	LMICs = 10Multiple = 2UMICs = 11	Quantitative = 16Qualitative = 1Mixed method = 6	Experimental = 7Quasi-experimental designs = 4Preexperimental = 2Other observational = 5Multiple = 5	Micro = 20Meso = 2Macro = 1	Pilot = 22Scale-up = 1	Adoption = 1Feasibility = 18Fidelity = 3Multiple = 1	10	Researchers = 10MOH = 5NGO = 1Providers = 6NC = 1	Behavior = 10Health outcomes = 8Behavior and health outcomes = 5	Demographic subgroup = 1Disease risk subgroup = 22
Diabetes	Glycemic control among people with diabetes	7	LMICs = 3UMICs = 2Multiple = 2	Quantitative = 5Mixed method = 2	Experimental = 2Quasi-experimental designs = 4Multiple = 1	Micro = 4Meso = 2Macro = 1	Pilot = 6Scale-up = 1	Adoption = 1Appropriateness = 2Feasibility = 2Multiple = 2	4	Researchers = 5MOH = 1NGO = 1	Behavior = 4Health outcomes = 2Behavior and health outcomes = 1	Demographic subgroup = 1Disease risk subgroup = 6
	Screening to prevent complications among people with diabetes	17	LMICs = 10UMICs = 7	Quantitative = 16Mixed method = 1	Experimental = 1Quasi-experimental designs = 1Preexperimental = 6Other observational = 8Multiple = 1	Micro = 16Meso = 1	Pilot = 1	Acceptability = 1Adoption = 2Feasibility = 10Multiple = 1Reach = 3	6	Researchers = 8MOH = 1NGO = 3Providers = 4NC = 1	Behavior = 5Health outcomes = 12	Disease risk subgroup = 16CHWs = 1
	Diabetes management	39	LMICs = 22UMICs = 14Multiple = 3	Quantitative = 33Qualitative = 1Mixed method = 5	Experimental = 8Quasi-experimental designs = 6Preexperimental = 9Other observational = 12Multiple = 4	Micro = 34Meso = 3Macro = 2	Pilot = 38Scale-up = 1	Acceptability Adoption = 1Appropriateness = 2Feasibility = 21Fidelity = 1Reach = 3Multiple = 4	16	Researchers = 22MOH = 6NGO = 4Providers = 6NC = 1	Behavior = 11Behavior and knowledge = 3Health outcomes = 20Behavior and health outcomes = 4Knowledge and health outcomes = 1	Demographic subgroup = 3Disease risk subgroup = 35CHWs = 1
	Influenza vaccination for people with diabetes	1	UMICs = 1	Quantitative = 1	Other observational = 1	Micro = 1	Pilot = 1	Adoption = 1	0	Researchers = 1	Health outcomes = 1	Disease risk subgroup = 1
Cancer	Breast cancer screening	9	LMICs = 5UMICs = 4	Quantitative = 9	Experimental = 1Quasi-experimental designs = 2Preexperimental = 2Other observational = 4	Micro = 6Macro = 9	Pilot = 7Scale-up = 1	Acceptability = 1Adoption = 1Feasibility = 4Implementation cost = 1Reach = 1Multiple = 1	7	Researchers = 4MOH = 4NC/NA = 1	Behavior = 4Health outcomes = 5	Demographic subgroup = 9
	Cervical cancer screening	93	LICs = 13LMICs = 42UMICs = 34Multiple = 4	Quantitative = 78Qualitative = 8Mixed method = 7	Experimental = 16Quasi-experimental designs = 4Preexperimental = 20Other observational = 47Multiple = 6	Micro = 78Meso = 10Macro = 5	Pilot = 78Scale-up = 12	Acceptability = 22Adoption = 23Feasibility = 22Implementation cost = 4Maintenance = 1Reach = 10Sustainability = 2Multiple = 9	40	Researchers = 56MOH = 13NGO = 10Providers = 2NC = 6NA = 6	Behavior = 30Health outcomes = 41Knowledge = 5Knowledge and behavior = 4Knowledge, behavior, health outcome = 2Knowledge and health outcomes = 3NC/NA = 8	Demographic subgroup = 80Disease risk subgroup = 2HCWs = 3CHWs = 2
	HPV vaccination for teen girls	5	LMICs = 2UMICs = 2Multiple = 1	Quantitative = 4Qualitative = 1	Preexperimental = 2Other observational = 3	Micro = 3Macro = 2	Pilot = 3Scale-up = 2	Adoption = 3Feasibility = 1Multiple = 1	1	Researchers = 3MOH = 1NGO = 1	Behavior = 2Health outcomes = 3	Demographic subgroup = 5
	Colorectal cancer screening	11	LMICs = 1UMICs = 10	Quantitative = 10Mixed method = 1	Experimental = 2Quasi-experimental designs = 1Preexperimental = 1Other observational = 6Multiple = 1	Micro = 7Meso = 4	Pilot = 9Scale-up = 2	Acceptability = 1Adoption = 4Feasibility = 3Implementation cost = 1Reach = 1Multiple = 1	4	Researchers = 7MOH = 4	Behavior = 5Health outcomes = 5Knowledge and behavior = 1	Demographic subgroup = 10Disease risk subgroup = 1
Chronic respiratory disease	Treatment of asthma	2	LMICs = 2	Mixed method = 2	Multiple = 2	Micro = 1Macro = 1	Pilot = 1Scale-up = 1	Acceptability = 1	1	Researchers = 1MOH = 1	Health outcomes = 2	Disease risk subgroup = 2

*Micro level refers to the point where the care providers interact with the patient; micro-level interventions aim to directly influence the performance of the staff or the operations of a facility [11,264]. Meso level refers to the level responsible for service areas/clinical programs providing care for a similar group of patients, typically part of a larger organization (e.g., subnational intervention targeting improvement of a network of facilities and communities) [11,264]. Macro level is the highest (strategic) level of the system, an umbrella including all intersecting areas, departments, providers, and staff (e.g., boards, healthcare network, integrated health system that includes several organizations); macro-level interventions are best able to directly tackle the social, political, economic, and organizational structures that shape a health system [11,264].

†Equity lens used if studies disaggregated by SES stratifiers (e.g., age, sex, education, income, and rural vs. urban) and/or targeted vulnerable population.

CHW, community health workers include ASHAs in India; CVD, cardiovascular disease; HCW, healthcare worker; HPV, human papilloma virus; LIC, low-income country; LMIC, lower middle-income country; MOH, Ministry of Health/Government; N, number of NCD interventions; NC/NA, not clear/not applicable; NCDs, noncommunicable disease; NGO, nongovernmental organization; UMIC, upper middle-income country.

Fig 2

Distribution of priority NCD prevention and control interventions by type of NCD and their risk factors (N = 265).

Fig 3

Distribution of studies per 1 million population by country of implementation.

We used country population size in 2020 (https://data.worldbank.org/indicator/SP.POP.TOTL) to standardized estimates expressed as number of studies per 1 million population. We used “rworldmap” package (https://cran.r-project.org/web/packages/rworldmap/rworldmap.pdf) available in R software to present these standardized estimates across countries where interventions were implemented. Country borders in this package are derived from Natural Earth data. Table E in S1 Appendix shows number of included studies per country.

Fig 4

Distribution of study countries, funding, and authorship (N = 222).

Fig 5

Growth of research over time (A) and distributions of NCD interventions by type (B). Fig 5A shows number of studies published each year (N = 222 studies); Fig 5B shows distributions by type of interventions (N = 265 NCD interventions evaluated in studied included in the review).

Fig 6

Distribution of implementation outcomes.

Most studies used quantitative methods, which accounted for 86%, whereas mixed methods and qualitative methods accounted for 9% and 5%, respectively (Table 2). The majority of studies used observational designs, with cross-sectional designs used in 45 studies. Among evaluations, preexperimental studies (such as pre-post without a comparison group or post-only) was the most frequently employed (n = 56 or 25% of all studies); experimental designs were used in a quarter of studies (n = 53 or 24% of all studies); quasi-experimental evaluation designs (such as pre-post comparison group or time series) were used in 15 papers (7% of all studies) (Fig 7). Study designs also appear to vary by NCD conditions targeted (Fig E in S1 Appendix). The sample size among included studies varied, ranging from 11 to 350,581, with median of 658. Most studies were standalone implementation studies (85%), with some variations by NCD conditions (Fig F in S1 Appendix). Hybrid implementation and effectiveness studies accounted only for 15%. Less than 5% of studies reported they were guided by widely known implementation science framework. Majority of studies were proof of concept or pilot versus scale-up studies (88% versus 12%), with variations by NCD conditions (Fig G in S1 Appendix). The level of health system targeted most often was micro level, accounting for 79% of studies, with variations by NCD conditions. The meso and macro levels of health systems were targeted by 14% and 7% of studies, respectively (Fig H in S1 Appendix). Approximately 42% of studies employed an equity lens—i.e., studies disaggregated by SES stratifiers (e.g., age, sex, education, income, and rural versus urban) and/or targeted vulnerable population.

Fig 7

Study designs.

A majority of studies (72%) reported funding, with international funding being the predominant source (Fig 4B). There seems to be some variations by NCD conditions (Figs I–K in S1 Appendix). For example, while 78% of studies focused on cervical cancer reported funding, of which 77% were from international sources, those focused on colorectal cancer and treatment of acute myocardial infarction received most of their funding from the countries where implementation research was conducted (Fig K in S1 Appendix). Majority of reported funding was provided by government/universities (43.6%), 35% reported multiple funders, 16% were foundations/NGOs, and 6% were private funders (e.g., pharmaceutical companies, professional associations) (Fig L in S1 Appendix). Approximately 62% of corresponding authors were from the country of implementation (Fig 4C); however, this varied by funding sources, with studies funded by international funders having the highest number of international corresponding authors.

Discussion

We conducted a systematic review of implementation research studies on NCD prevention and control strategies in LMICs published between 1990 and 2020. We focused our analysis on WHO-recommended NCD interventions carried out by the health system rather than through policy, legislation, or public health approaches [6,7]. These studies therefore represent the state of the implementation science in prevention and control of NCDs by health systems in the countries bearing the bulk of disease burden from noncommunicable conditions.

Of the 222 implementation science studies included in this review, 94% were conducted in middle-income countries (evenly split between lower- and upper-middle) and 6% in LICs. UMICs were slightly overrepresented compared to their share of the LMIC population (approximately 40%). Only 8 of the studies were multicountry studies, suggesting that cross-national generalizability is not the primary motivation for this type of research. India and China, with 43% of the population of LMICs, comprised one-third of the studies. South Africa, Brazil, Iran, Kenya, and Nigeria, were well represented, each contributing more than 3% of the research.

The studies described 265 different NCD interventions, ranging from screening to prevention to treatment and palliation. Conditions studied varied substantially by region. All 13 of the interventions studied in LICs were for cervical cancer screening. In low-middle income countries, cervical cancer accounted for 37%, diabetes for 29%, and hypertension for 8% of interventions. There was a larger variety of conditions studied in UMICs: while cervical cancer and diabetes comprised half the studies, hypertension, myocardial infarction, colorectal cancer, other cardiovascular diseases, and unhealthy diet each comprised more than 5% of studies. The 2 countries with the largest research output and populations, China and India, differed substantially in focus. In India over 70% of studies were on 2 conditions: diabetes (51%) and cervical cancer (19%), whereas the research was more evenly distributed across the NCDs in China.

Half of all studied interventions in this review evaluated screening for disease, nearly 40% treatment and 12% prevention. Over 70% of all screening studies were for cervical cancer, with less research on other conditions for which screening can be cost effective, such as diabetes, colorectal cancer, and breast cancer. Primary and secondary prevention can reduce incidence of disease and forestall disease progression and disability. We found that only 31 (12%) of the studied interventions addressed prevention with nearly 80% of these tackling prevention of the NCD behavior risk factors (e.g., tobacco use, inactivity, unhealthy diet). Less than 10% of the interventions evaluated in this review focused on management of hypertension (the leading metabolic risk factor worldwide, accounting for approximately 19% of global deaths) [246]. This suggests a substantial implementation research gap in secondary prevention, a critical function of primary care and other levels of health systems. Primary care services such as hypertension management and glucose control play a major role in reducing mortality, thus insufficient research on their optimal implementation is a major missed opportunity. Recent work shows that treatment and control rates for hypertension were below 25% and 10%, respectively, in many countries in South Asia and sub-Saharan Africa. These countries also showed the slowest rates of improvement from 1990 [247].

The preponderance of interventions studied was in pilot phase, with fewer than 15% studying large-scale implementation. Along the same lines, feasibility and adoption were the most studied implementation outcomes, suggesting the research is focused on introduction of new approaches. While proof of concept studies is vital with new implementation strategies, arguably WHO-proposed interventions are well established and evidence on (clinical) effectiveness abound. To provide useful guidance to health system planners and realize population health gains, there needs to be a greater investment in large-scale NCD implementation research to promote sustainability of evidence-based interventions. To best scale scarce research resources and accelerate impact, countries could join regional consortia to study interventions and undertake factorial designs that compare locally adapted implementation approaches.

Over three-quarters of the studies were situated at the micro level of the health system—targeting patient, provider, or clinic levels. Nearly 1 in 5 tested a new technology, despite evidence that technology adoption without substantial integration into policies, data, and workflows is typically ineffective in transforming care [248,249]. Education was another common target featuring in 3 of 10 studies; researchers accounted for the majority of the actors. While micro-level approaches are the most “researchable,” as they are easiest to implement and analyze; positive results are difficult to scale and sustain in the absence of systemic health system change. The Lancet Commission on High Quality Health Systems notes that high-quality care results from structures that align system aims and policies with strong governance, management, and appropriately trained workforce [250]. In this context, micro-level innovation cannot raise quality system wide and is only effective if undertaken as part of a learning health system that can determine whether it offers sufficient benefit over current practice in complexity, cost, and health benefit, and if so, how to best integrate into the health system [251].

Nearly 9 in 10 studies were stand-alone implementation research. This also points to an opportunity to add implementation research to ongoing effectiveness trials. Integrated or hybrid effectiveness-implementation studies are increasingly being used in high-income countries to shed light on both the outcome and extent and quality of service/program delivery [252]. Notably, fewer than 5% of studies cited use of an implementation science framework consistent with prior research showing that the use of implementation science framework is substantially lower in LMICs compared with high-income countries [253]. The use of a tested conceptual framework can improve the rigor of the research and promote comparability of results. Of the studies that reported a funding source, 60% was from international sources, 33% from the country of the research, and the remaining from both local and international sources. This reflects the low spending for health research and especially for health systems and implementation research in LMICs. The lack of domestic support is unlikely to be offset by global funding going forward; a recent analysis showed that NCDs were under prioritized in bilateral agency portfolios relative to their health impacts [254]. Over 40% of development assistance for health in LICs for NCDs came from NGOs and philanthropies, which are less inclined to support research than operations [254]. Indeed, we found that only 16% of studies with funding information reviewed were supported by philanthropies or NGOs, while the other remaining studies reported funding sources from government, private, and/or multiple sources.

Scarcity of funding for research is a key constraint to needed implementation research for NCDs. While there are proposals for coordinating and increasing global support, it is unrealistic to expect this to meet the scale of needed research without a substantial increase in countries’ investment in research [255]. Such an investment is likely to pay off in better health and higher quality, more efficient service delivery [256]. To make best use of research funds, implementation science should strive to be as generalizable as possible—at minimum at a regional level where health systems share similarities. International and regional institutions can play an important role in supporting research consortia and partnerships to promote efficiency of and accelerate the pace of research and, ultimately its uptake into routine care at scale.

Over 50 of the 222 included studies used an experimental research design. While this is the strongest design to yield causal inference, it is not always feasible to implement. Quasi-experimental designs, such as pretest, posttest comparison group designs, and interrupted time series, which can offer robust information were used in only 15 studies. Preexperimental designs that do not include a comparison group or tracking over time, comprised nearly a quarter of the studies. These designs have very low internal validity and should generally be avoided. The remainder of the studies used cross-sectional descriptions, cohort studies, and qualitative research or multiple study types. Given the disproportionate health harms of NCDs among the poor and other vulnerable groups within countries, disaggregated or stratified analysis is crucial. Forty percent of the assessed studies included stratification by age, sex, education, or urbanicity. Going forward, greater use of quasi-experimental designs, hybrid implementation studies and mixed methods approaches, would benefit the field. An expanded focus on equity of implementation outcomes is also needed.

Strengths and limitations

Our study had several strengths, notably the extensive scope for the search that covered LMICs, a wide range of outcomes and study types, and a large contingent of conditions and health services. We had no language restrictions permitting a comprehensive assessment of the published literature. The review also had several limitations. We focused on WHO-recommended interventions, which at present do not include guidance for some prevalent conditions such as mental health problems and kidney disease [6,12]. Mental health is a major contributor to the global burden of disease and future work should assess the implementation science for the growing range of mental health interventions that appear to be effective in lower-income settings [257,258]. The studies we assessed used differing definitions of implementation outcomes (e.g., acceptability was measured in some studies by self-report and in others by behavior change). This limits direct comparison of study outcomes. Greater use of implementation science frameworks can promote coherence in the research approaches and terminology used to the benefit of end users. Similarly, given the implementation strategies were not specified well enough in the included studies, we elected to focus on actors, action target, and recipients in our description of implementation strategies. Clearly, reporting empirical implementation studies using existing framework to describe implementation strategies would help bolster uptake of implementation research in NCDs.

We also did not search the gray literature and as such, some relevant studies may have been missed. However, studies in gray literature that were not peer reviewed would have not have been eligible for inclusion in this review. Despite using rigorous search strategies without language restrictions, studies published in journals not indexed in MEDLINE and EMBASE were not captured [259-263]. Given the focus on this review and the heterogeneity in aims and methodologies of included studies, risk of bias assessment to understand how effect size may have been compromised by bias is not applicable. As such, we only commented on the distribution of research designs and discussed about stronger/weaker designs. Lastly, we reported year of publication and not time of when study/implementation was conducted.

Conclusions

High-quality implementation science can play a key role in informing effective delivery of health system interventions to mitigate the burden of NCDs and avoiding expensive mistakes. While implementation research on priority NCDs has grown substantially, from under 10 studies per year in early 2000s to 51 studies in 2020, this is still vastly incommensurate with the health importance of the topic. Further, the concentration of studies in a few geographies and a few health areas, such as cervical cancer, highlights the dearth of research for other key conditions. We found a major gap in research on secondary prevention, i.e., management of risk factors or early disease to prevent disease progression and premature death. Research on ways in which health systems can be strengthened, including primary care levels, to provide optimal care for NCDs is critically needed. Future studies should use implementation science frameworks, and, when testing interventions, strong research designs with strong internal validity, including well-designed quasi-experimental studies. Opportunities exist for adding implementation science studies to planned effectiveness research.

PRISMA 2020 checklist.

(DOCX)

Click here for additional data file.

Appendix tables and figures.

Table A in S1 Appendix. Interventions provided within health systems. Table B in S1 Appendix. Sample of the search strategy used in the MEDLINE database. Table C in S1 Appendix. List of low- and middle-income countries. Table D in S1 Appendix. Data extraction tool. Table E in S1 Appendix. Distribution of studies by countries where they were implemented. Fig A in S1 Appendix. Variation of conditions evaluated by income group. Fig B in S1 Appendix. Priority NCD interventions (n = 265) identified in 222 studies included in the review. Fig C in S1 Appendix. Distribution of included studies by NCD. Fig D in S1 Appendix. Distribution of intervention type by income group. Fig E in S1 Appendix. Distributions by research designs. Fig F in S1 Appendix. Distributions by standalone implementation studies vs. embedded or hybrid effectiveness-implementation studies. Fig G in S1 Appendix. Distributions by pilot vs. scale-up project. Fig H in S1 Appendix. Variation by level of health system. Fig I in S1 Appendix. Studies that reported funding (vs. those that did not) by NCD conditions. Fig J in S1 Appendix. Distributions by funding type. Fig K in S1 Appendix. Distribution of funding sources by NCDs and their risk factors. Fig L in S1 Appendix. Types of reported funding sources (N = 222 included studies).

(DOCX)

Click here for additional data file.

10 Dec 2021

Dear Dr Hategeka,

Thank you for submitting your manuscript entitled "Systematic Review of Implementation Research on Non-Communicable Disease Care in Low- and Middle-Income Countries" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by Dec 14 2021 11:59PM.

Login to Editorial Manager here: https://www.editorialmanager.com/pmedicine

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Beryne Odeny

PLOS Medicine

22 Dec 2021

Submitted filename: Response to the PLOS Medicine Staff.docx

Click here for additional data file.

28 Jan 2022

Dear Dr. Hategeka,

Thank you very much for submitting your manuscript "Implementation Research on Non-Communicable Disease Care in Low- and Middle-Income Countries: A Systematic Review" (PMEDICINE-D-21-05057R1) for consideration at PLOS Medicine.

Your paper was evaluated by a senior editor and discussed among all the editors here. It was also sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by Feb 18 2022 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Beryne Odeny,

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

1) Please add this statement to the manuscript's Competing Interests: "Margaret Kruk is an Academic Editor on PLOS Medicine's editorial board."

2) Abstract:

a) Please report your abstract according to PRISMA for abstracts, following the PLOS Medicine abstract structure (Background, Methods and Findings, Conclusions) http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.1001419 .

b) Please combine the “Methods” and “Findings” sections into one “Methods and findings” section

c) Please ensure that all numbers presented in the abstract are present and identical to numbers presented in the main manuscript text.

d) Please summarize the synthesis/appraisal methods

3) Please remove the “Research in Context” section

4) At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary

5) Please submit a completed PRISMA checklist. We understand that not all items on the checklist will be directly relevant, and these can be marked as ‘not applicable’.

a) The PRISMA guidelines provided at the EQUATOR site http://www.equator-network.org/reporting-guidelines/prisma/

b) When completing the checklist, please use section and paragraph numbers, rather than page numbers.

6) Please rename figure 1 to “PRISMA flow chart”

7) In line with PLOS Medicine’s guidelines, please update your search to the present time and provide the beginning and end dates of your search.

8) References:

a) Please select the PLOS Medicine reference style in your citation manager. In-text reference call outs should be presented as follows noting the absence of spaces within the square brackets, e.g., "... countries [1,2]."

b) Please ensure that journal name abbreviations consistently match those found in the National Center for Biotechnology Information (NCBI) databases. https://journals.plos.org/plosmedicine/s/submission-guidelines#loc-references.

c) Please ensure six names appear before et al. e.g., check refs #26-32, 34-36, 38 and so forth

Comments from the reviewers:

Reviewer #1: See attachment

Michael Dewey

Reviewer #2: Reviewer Comments

Title: Systematic Review of Implementation Research on Non-Communicable Disease Care

in Low- and Middle-Income Countries

Authors: Hategeka C, et al.

Overview

The investigators conducted a systematic review of evidence on the current state of implementation research (IR) for NCD prevention and control in low-and middle-income countries (LMICs) using MEDLINE and EMBASE databases from 1990 through 2020. They used standard search terms for this topic that are detailed in their documentation. Among more than 9600 potential studies they found 222 eligible studies from 63 countries. Most studies were on cervical cancer and were for proof of concept or pilots targeted at a micro-level of the healthcare system. Most studies used quantitative methods and weak study designs were common. Study publications increase dramatically over their study timeframe. Major gaps in IR were noted and that published studies were mostly funded by international sources.

General Comments

This is a very well written study that is very important and timely. IR is currently vastly underused in LMICs, and the authors highlight this challenge and the need for more efforts and resources. While they do show convincingly that the number of IR studies annually has increased over their 30-year study timeframe, efforts in this area fall short of what is needed.

There are several areas where the manuscript could be improved. The results seem to be in both the results section some are also in the discussion section. There are many figures and tables and some of the findings could be moved into the text with a concise description and allowing for deletion of some figures/tables. Better description of the inclusion/exclusion criteria would be helpful. A brief clear description of the working definition of screening, treatment and prevention would be useful.

Some discussion on why so few studies were multinational would be helpful. Also, a brief description of micro, meso and macro health system interventions would be useful. In the discussion, using high income country comparable statistics might be useful to give the context of what is currently the situation (e.g., use of implementation science framework for studies in HIC vs LMIC, etc.). While not the ideal comparator is could help under the challenges in the LMIC context. In the discussion/conclusion, it would be helpful to present some of the challenges and barriers conducting IR in LMICs. This could be very insightful. Finally, a better description of why descriptive epidemiological studies were excluded is needed. These may contain barriers and facilitators to implementation and be key in designing IR intervention strategies.

Specific Comments

Funding source section.

While the statement is made that the funders have no involvement in the study, it does not state who the funders were. Suggest including them.

Page 8 first para

…"country of implementation" is a bit unclear. Suggest making it "county where the IR is being conducted" - it seems that is what you are communicating.

Figure 1.

More description of the exclusion categories is needed. For example, intervention, and full text are not entirely clear.

Figure 2

How were studies that had multiple interventions in this graphic? Were they included in all the categories that had interventions?

Figure 3.

This plot is driven by county size. Consider a metric that uses a standardization method such as studies per 100,000 population. This may allow for better comparisons, etc.

Figure 5.

Seems that A and B should have the same total numbers but in B 2020 is lower than 2020 in A. Please explain/reconcile.

Figure 7.

Same issue as for Fig 2. How were studies with multiple interventions handled. Some studies may have a N >1. Please clarify.

Table 1.

See above comment about exclusion of epidemiology descriptive studies. They may have barriers and facilitates for IR intervention development.

Appendix Table 3.

Should have a date on it because LMIC countries change their status over time (e.g., move from LIC to LMIC, etc.)

Appendix Figure 1.

Diabetes is misspelled

Appendix Figure 5.

See above comments. Macro, Meso, and Micro need short descriptions.

Appendix Figure 7.

The legend is off the page.

Appendix Figure 8.

Legend is confusing as a standalone. Suggest: Funding, no funding (at least none noted in the published report). However, for the study to happen, it must have been funded at some level - so 'no funding" may not work. More detail would be interesting.

Reviewer #3: Thank you for this interesting and useful review of the literature focusing on implementation research on 33 WHO recommended interventions to prevent and control non-communicable diseases in low- and middle-income countries.

General comments

Overall, the study is well written and easy to follow. I have a series of more specific suggestions and comments below, but my main concern is that you only present descriptive statistics regarding some key characteristics of the identified studies, rather than the narrative synthesis which you planned to do according to the study protocol in PROSPERO. Such a narrative synthesis could have provided the reader with a richer understanding of what included studies found and other aspects beyond the narrower quantitative approach used. If you can, including a narrative synthesis would likely be very interesting. If this is not feasible, you should at least comment on this departure from the study protocol and throughout the paper more clearly state the narrow quantitative scope of the study.

Another key aspect which is missing from the paper is an assessment of the quality and usefulness of the included studies. Indeed, you stress the importance of good quality implementation research in several places in the manuscript, yet it is difficult from this study to say anything about the current quality of implementation research on NCD interventions in LMICs.

Specific comments

-- Title

Since the study focuses on NCD prevention and control interventions and not only care, you could consider replacing "Care" with "prevention and control interventions" in the manuscript title: "Implementation Research on Non-Communicable Disease Prevention and Control Interventions in Low- and Middle-Income Countries: A Systematic Review"

-- Abstract

I would not mention the protocol in the abstract.

"We synthesize extracted data narratively using descriptive statistics" - generally narrative synthesis refers to "the use of words and text to summarise and explain the findings of the synthesis", whereas descriptive statistics is a quantitative approach to describe or summarize the characteristics of a sample or data set. Actually, in this study you only present descriptive statistics, no narrative summary or synthesis. Please revise the abstract accordingly.

It would be useful if you could mention something about inclusion and exclusion criteria in methods.

"…approximately similar to their proportion of global population…" while this is a nice idea, it does not make sense to make this sort of comparison since you are missing the high-income countries in your review and they would very likely distort these proportions. Also, your assessment isn't quite correct since LICs are home to about 9% of the global population and about 73% live in MICs.

"Slightly more studies used the weakest study design, pre-experimental, than the strongest design, experimental (25% versus 24%)" - in my view, it would be more correct to state that these two categories were (almost exactly) equal as the difference is something like 2 studies out of 222.

"Despite growth in implementation research on NCDs in LMICs" - this statement is not supported by the findings in the abstract, suggest adding a line supporting this.

"Future studies should prioritize implementation at scale". One could argue that studying interventions at the pilot stage should be prioritized in order to assess whether to take an intervention to scale in the first place. Is there a risk that advocating for a priority on studying implementation at scale could lead to less well studied interventions going to scale?

"stronger internal validity, be more conceptually-driven and use mixed- methods to understand mechanisms" - while interesting, these conclusions are not supported by the findings stated in the abstract. Note that it should be possible to read the abstract without reading the full manuscript.

"To maximize impact of the research under limited resources, adding implementation science outcomes to effectiveness research and regional collaborations are promising." Could you clarify what you mean by this statement, and consider whether this conclusion is supported by the findings of the present study?

-- Research in context

The text provided under "Added value of this study" does not seem to correspond to this heading. I would suggest beginning the paragraph with something like "This study provides the first comprehensive review of implementation research on NCD interventions…"

The text provided under "Implications of all the available evidence" does not seem to be supported by the findings of the present study. Please review this paragraph.

-- Introduction

In the first paragraph, you only mention NCD mortality. I would encourage you to also mention the significant NCD morbidity and its consequences.

You mention COVID-19 and the interlinkages between NCD morbidity and COVID-19. You could consider also making a similar link to other infectious diseases such as TB and HIV to drive home the point that NCD prevention and control should not be regarded as a vertical issue separated from other health issues.

You write that "…the evidence for the clinical effectiveness of most NCD prevention and treatment interventions is well established…" - would you agree with Isaranuwatchai et al. (BMJ 2020;368:m141) who argue that the clinical effectiveness of some NCD prevention and treatment interventions may indeed be contestable or even wasted in some contexts?

Please consider explicitly stating the study aim and/or research question of the study.

Given the limited scope of this study, consider rephrasing "In this systematic review, we synthesize evidence on the current state of implementation research" to better reflect that you present descriptive statistics of key characteristics of peer-reviewed publications on implementation research on 33 specific interventions in LMICs in 1990-2020.

-- Methods

According to the information registered on the PROSPERO database, the study protocol was technically not pre-specified as the start date (Dec 1, 2020) was seven months before the protocol was registered (July 30, 2021), and the anticipated completion date was only one day later (July 31, 2021).

Could you clarify in the methods how you selected the included interventions and specifically how you define "relevant" interventions and "within health systems"? For example, in Table 2 / Appendix table 1 you list "mass media campaigns that educate the public about the harms of smoking/tobacco use and second hand smoke" as an intervention within the health system, whereas it seems like you have excluded other interventions such as taxation and plain packaging.

"Low- and middle-income countries as defined by the World Bank." - please specify which version you use as the classification is modified each year and the list has changed a fair bit since 1990. For example, in 1990 both China and India were classified as low-income countries.

Throughout, I would recommend making the reference to the updated Appendix 3 of the WHO Global NCD Action Plan 2013-2020 more explicit to avoid the perception that you came up with your own list of interventions.

The list of data elements you extracted, presented in appendix table 4, plays a central role in defining the focus of your study. Please describe how you arrived at the 29 included parameters.

The following sentence is redundant since the information is provided in table 1: "Non-empirical/primary research including reviews meta-analyses, editorials, commentaries, letters to editors, opinion papers, newspaper and protocols, are not eligible for inclusion."

As noted in the abstract: As far as I can see, you do not present any narrative summary or synthesis of findings, only some basic descriptive statistics.

The heading "Risk of bias assessment" could be removed from methods since it was not done, unless it is explicitly required by editors. The corresponding paragraph could be moved to the limitations section.

Table 1. This table is unnecessarily unwieldy. Specifically I would recommend removing the row for "population" as your review focuses on interventions and not populations (that non-humans are excluded can be left implicit in my opinion), and significantly shorten the study design section - it is not necessary to list all existing study designs here.

-- Results

Perhaps I am missing something, but I would normally expect there to be a table with the full list of included studies and their characteristics (either in manuscript or in appendix). Currently, there is just a reference to Table 1 (which should not be referenced here as it belongs to methods), followed by references "20-241". In my view, this is not a sufficient way to list the included studies.

I would recommend using the same number of decimal points throughout (you now mix between 1 and 0 decimal points).

You refer to a grand total of 12 figures in the Appendix, all of which are simple bar or pie charts. In my opinion, this information would be far more easily digested if you could present it as one or two tables instead.

According to table 2 you have only 33 eligible interventions in this study - how could you find 265 interventions? Per the study design, you should only be able to find a maximum of 33 interventions. Could you clarify this? (According to Appendix figure 1, you found studies covering 32 of the 33 eligible interventions, and in total there were 265 instances across the 32 interventions).

Appendix figure 1. The figure says "Fiabetes management", please review.

How do figure 2 and appendix figure 1 relate? The numbers seem not to match.

Appendix table 5 should have a column with country income category. The column entitled "proportion" should read "percentage".

Sorry to make an editorial comment, but please make sure the order of figures and tables are correct (for example, Appendix figure 2 appears in the text before Appendix figure 1).

You state that "The NCD conditions targeted varied by countries" and refer to Appendix Figure 3, but the figure only shows how the number of interventions targeting each condition (though you write number of conditions evaluated) varied by country income category. Please revise the text and figure.

Could you clarify what you mean by "The NCD interventions varied by conditions and type" - in what way did interventions vary? Perhaps you mean that the number of identified interventions varied?

When you state "The intervention focus appears to vary by countries" you again seem to be referring to the country income category rather than country.

The three types of intervention focus brought up in results (screening, treatment and prevention) do not appear to be presented neither in the methods nor in the appendix - where Appendix table 4, row 22. "intervention type" refers back to Appendix table 1, where in column "category of interventions" you only have primary, secondary or tertiary prevention. Please make sure that all methods used are presented in the methods and that the same terminology is used throughout.

Please include a definition and reference for what you mean by "level of health system" (micro, meso, macro) in the methods as this is not clear to the reader.

Each time you write that study characteristics varied by income category or condition etc. I wonder whether you did any simple statistical test to assess whether there were significant differences, or did you decide against this?

I would suggest mentioning the equity lens in the methods since you bring it up in the results (even though it is presented in Appendix table 4).

Figure 5A presents the number of publications by year, but it generally takes years from research to publication - do you have any sense of when research was carried out, not only when it was published?

-- Discussion

I am not sure I agree with this statement: "These studies therefore represent the state of the science today on how to scale up the response of the health system to the growing burden of NCDs in the countries bearing the bulk of disease burden from non-communicable conditions." - nowhere in your study design do you mention a focus on "how to scale up the response of the health system", or is this implicit somehow?

Please rephrase as discussed above: "The studies described 265 different NCD interventions, ranging from screening to prevention to treatment and palliation."

The comparison between number of studies and disease prevalence is interesting but I would recommend adding it to results (and methods) rather than introducing this analysis in the discussion.

Some of the statements in the discussion appear to be new information, i.e. "We found that only 31 of the studied interventions addressed prevention with nearly 80% of these tackling primary prevention (e.g., tobacco use, inactivity, unhealthy diet)." - could you please make sure that you do not introduce new information in the discussion?

You write: "arguably WHO- proposed interventions are well established and a range of implementation models abound" - please see my earlier comment about the fact that the list of interventions has been questioned in its own right. I also note that you contradict yourself when you write that "a range of implementation models abound", while in the introduction you noted that "care delivery models and means of scaling these up to entire populations in need in heterogeneous and resource-constrained health systems are not".

"the micro-level of the health system— targeting patient, provider, or clinic levels", this is the first time in tha manuscript that you explain what you mean by the micro level, see comment above.

"Nearly one in five tested a new technology, despite evidence that technology adoption without substantial integration into policies, data, and workflows is typically ineffective in transforming care" - are you here implying that this was not the case in the included studies?

"The lack of domestic support is unlikely to be offset by global funding going forward; a recent analysis showed that NCDs were under prioritized in bilateral agency portfolios relative to their health impacts." - I would argue that this situation is changing, besides, it may be more useful to think about this in terms of health systems strengthening than as a vertical NCD silo. You may want to touch on that aspect in the discussion.

You write "We had no language restrictions permitting a comprehensive assessment of the published literature" - did you do any searches in other languages? Did you include studies in other languages?

I would suggest adding a few sentences to the limitations about the fact that you did not assess quality or usefulness of the included studies. (Indeed, you begin the conclusion by stating that "High quality implementation science can play a key role in informing effective delivery of health system interventions", yet your study only quantifies the number of studies, not their quality.)

Limitations should also, importantly, discuss the fact that there may be significant implementation research conducted which is never published which may introduce important selection bias in your study.

The paragraph starting with "Over 50 of the 222 included studies used an experimental research design." seems misplaced after the "strength and limitations" section.

-- Conclusion

You should generally avoid adding new arguments and references in the conclusion, but rather conclude what you found, relating back to your study aim. Your current conclusion reads more like a conclusion about the usefulness and need for implementation more broadly, rather than about your study findings. Several of the conclusions are not supported by the study findings and should be brought up in the discussion rather than the conclusion. Please revise the conclusion accordingly.

Any attachments provided with reviews can be seen via the following link:

[LINK]

Submitted filename: hategeka.pdf

Click here for additional data file.

28 Feb 2022

Submitted filename: Hategeka et al. Responses to EditorsReviewers FEB25.docx

Click here for additional data file.

29 Mar 2022

Dear Dr. Hategeka,

Thank you very much for submitting your manuscript "Implementation Research on Non-Communicable Disease Prevention and Control Interventions in Low- and Middle-Income Countries: A Systematic Review" (PMEDICINE-D-21-05057R2) for consideration at PLOS Medicine.

Your paper was evaluated by a senior editor and discussed among all the editors here. It was also discussed with an academic editor with relevant expertise, and sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by Apr 19 2022 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Beryne Odeny,

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the academic editor:

This is an interesting study, but I think two things are missing and could make it much better.

First, there is a notable lack of attention in systematically capturing categorizing the strategies. While the categorization of strategies is a moving target in implementation science, it is nevertheless a crucial "ingredient" in implementation science and our limited ability to describe the strategy (which is the exposure or treatment if you will in this field) limits what we learn from it. There are many different ways of classifying strategies, that range from the original Proctor Actor-Action etc., to ERIC to Cochrane EPOC to Waltz et al. - none of them are perfect but they can still be useful. At a minimum, who in the health system is carrying out the strategy, what are the actions, and who are the actions meant to target are critical. In truth, some of the strategies are going to be difficult to categorize because the primary sources do not say what they are, but it would nevertheless be of interest say how many did or did not. The authors are undoubtedly familiar with both the challenges and the importance of describing strategies used in these studies, so whatever direction they take, some discussion and justification of it in the discussion would be helpful.

Some additional analysis of the implementation outcomes might be useful, with more description on where in the health system (at the patient, the hcw, the organization, the system) etc., the strategies targeted. How has this changed over time? does it differ by disease condition? what about by donor? Likewise this could be done with the strategies as well, but understanding what kind of targets implementation science addressing NCD's seek would be helpful. What I would be concerned about in general is the phenomenon where clinical research and patient level outcomes in clinical medicine drive outcome selection and measurement, and even though much of the action is above the level of the patient, the outcomes - even those considered implementation outcomes - are reported at the level of the patient, and if so would be of limited informativeness.

Comments from the reviewers:

Reviewer #1: The authors have met most of my points especially about the figures although I still think Figure 4 is not really a good use of ink and space.

I thank the authors for the offprint about comparison of databases but I do not draw the same optimistic message from it as they do. Unfortunately it does not give us the breakdown of the intersections of the coverage so we cannot tell if searching a new database would give us additional references. Just because Embase has, say 10%, and AIM 5% does not imply the 5% are a proper subset of the 10%. They might all be different. I still think the search is rather limited.

Michael Dewey

Reviewer #3: Thank you for your thorough and thoughtful responses to reviewers' and editors' comments. As this is a re-review, I present my remaining concerns below.

You state that "We have already clarified in our methods section and in the protocol that narrative descriptive analysis that would be used in this review refers to descriptive statistics including summary statistics of type of interventions, study designs, and implementation outcomes". My point is that you cannot do descriptive statistics and call it narrative synthesis as those are two different things. Please revise the manuscript accordingly, possibly including an explanation why you did not do a narrative synthesis as you had planned to do according to the study protocol.

You state that "Table 2 shows a list of unique interventions that were eligible for inclusion. In the results, we report total number (instead of unique type of interventions) of NCD interventions evaluated in 222 included studies. We found that 265 interventions were evaluated across 222 included studies, meaning that there were some studies that evaluated more than one intervention.". I do not think you have addressed the problem I raised regarding the terminology used, namely that you use the word "intervention" to mean both the 33 priority NCD interventions included in this study and the 265 studied instances of those interventions being used. This is particularly confusing since you actually found 222 studies covering 32 of the 33 interventions. I suggest you consider an alternative terminology which separates between the 33 priority interventions and the 265 instances of interventions being used.

Regarding your comment that "We reported year of publication and not time of when study/implementation was conducted." I would suggest you make a comment about this in the limitations.

Any attachments provided with reviews can be seen via the following link:

[LINK]

9 May 2022

Submitted filename: Responses to editors and reviewers May 02.docx

Click here for additional data file.

1 Jun 2022

Dear Dr. Hategeka,

Thank you very much for re-submitting your manuscript "Implementation Research on Non-Communicable Disease Prevention and Control Interventions in Low- and Middle-Income Countries: A Systematic Review" (PMEDICINE-D-21-05057R3) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Jun 08 2022 11:59PM.

Sincerely,

Beryne Odeny,

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

1) Figure 3 - At the top of each column, please indicate the meaning of numbers in each cell/ row (e.g., “=5”). If this is the N of studies, please clearly indicate this at the top of each column.

2) References

a) Please ensure that journal name abbreviations consistently match those found in the National Center for Biotechnology Information (NCBI) databases. https://journals.plos.org/plosmedicine/s/submission-guidelines#loc-references.

b) Please ensure that an access date is provided for all references with weblinks.

Comments from Reviewers:

Any attachments provided with reviews can be seen via the following link:

[LINK]

2 Jun 2022

Submitted filename: Responses to Requests from Editors JUN01.docx

Click here for additional data file.

16 Jun 2022

Dear Dr. Hategeka,

Thank you very much for re-submitting your manuscript "Implementation Research on Non-Communicable Disease Prevention and Control Interventions in Low- and Middle-Income Countries: A Systematic Review" (PMEDICINE-D-21-05057R4) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Jun 23 2022 11:59PM.

Sincerely,

Beryne Odeny,

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

- In the abstract, methods and discussion sections, please clearly mention that the implementation strategies are not specified well enough. This should be highlighted as one of the limitations.

Any attachments provided with reviews can be seen via the following link:

[LINK]

18 Jun 2022

Submitted filename: Responses to Requests from Editors.docx

Click here for additional data file.

21 Jun 2022

Dear Dr Hategeka,

On behalf of my colleagues and the Academic Editor, Dr Elvin Hsing Geng, I am pleased to inform you that we have agreed to publish your manuscript "Implementation Research on Non-Communicable Disease Prevention and Control Interventions in Low- and Middle-Income Countries: A Systematic Review" (PMEDICINE-D-21-05057R5) in PLOS Medicine.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Beryne Odeny

PLOS Medicine