YPMED-04091; No of Pages 12 Preventive Medicine xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed

Effectiveness of physical activity promotion interventions in primary care: A review of reviews☆ Alvaro Sanchez ⁎, Paola Bully, Catalina Martinez, Gonzalo Grandes Primary Care Research Unit of Bizkaia, Basque Health Service-Osakidetza, Bilbao, Spain

a r t i c l e

i n f o

Available online xxxx Keywords: Physical activity Health promotion Primary care Review

a b s t r a c t Objective. The present review aims to summarize the evidence about the effectiveness of physical activity (PA) promotion interventions in primary care (PC) and the intervention or sample characteristics associated with greater effectiveness. Methods. MEDLINE, EMBASE, and Cochrane Library were searched to identify systematic reviews and meta-analyses published from 2002 to 2012 that assessed the effectiveness of PA-promoting interventions in PC. Information was extracted and recorded about each of the selected studies and their reported results. Methodological and evidence quality was independently rated by two reviewers using the nine-item OQAQ scale and the SIGN classification system. Results. Ten of the 1664 articles identified met the inclusion criteria: five meta-analyses, three systematic reviews, and two literature reviews. Overall, PA promotion interventions in PC showed a small to moderate positive effect on increasing PA levels. Better results were obtained by interventions including multiple behavioral change techniques and those targeted to insufficiently active patients. No clear associations were found regarding intervention intensity or sample characteristics. Conclusion. Although several high-quality reviews provided clear evidence of small but positive effects of PA intervention in PC settings, evidence of specific strategies and sample characteristics associated with greater effectiveness is still needed to enhance the implementation of interventions under routine clinical conditions. © 2014 Elsevier Inc. All rights reserved.

Background The numerous health benefits of regular physical activity (PA) are well known. Accordingly, it is recommended that adults perform at least 150 min/week of moderate-intensity PA, 75 min/week of vigorous PA, or a combination of moderate and vigorous PA (Haskell et al., 2007). However, a majority of the population in developed countries does not follow these recommendations, making PA promotion a public health priority (Tucker et al., 2011; Hallal et al., 2012). Primary care (PC) practitioners can play a key role in promoting PA and improving population health in developed countries because of the ongoing care they provide to a large sector of the population (Estabrooks et al., 2003). It is estimated that up to 80% of adults in these countries visit their general practitioner (GP) at least once a year ☆ Sources of support: The project has received funding from a Network for Prevention and Health Promotion in Primary Care (redIAPP, RD12/0005) grant, research project grants (PI13/00573, PI12/02635, PI12/01914and PS09/01461) from the Instituto de Salud Carlos III (Institute of Health Carlos III) of the Ministry of Economy and Competitiveness (Spain), co-financed with European Union ERDF funds, and Health Department of the Basque Government (EXP: 2009111072 and 2007111009). ⁎ Corresponding author at: Unidad de Investigación de Atención Primaria, Osakidetza, Luis Power 18, 4ª planta, E-48014 Bilbao, Spain. Fax: +34 946006639. E-mail address: [email protected] (A. Sanchez).

(van Doorslaer et al., 2006). Until recently, evidence about the effectiveness of interventions promoting PA in routine PC practice, especially in the long term, has been considered inconclusive (Foster et al., 2005; Muller-Riemenschneider et al., 2008). Newer studies have concluded in favor of PA interventions in the primary care setting, and recent meta-analyses indicate that the evidence appears to be shifting in this direction (Lin et al., 2010; Orrow et al., 2012; Hillsdon, 2013). The high prevalence of inactivity in the population and the many obstacles faced by PC professionals in a setting characterized by work overload and a shortage of time and specialized training (Estabrooks and Glasgow, 2006; Grandes et al., 2008) combine to support the need for clear evidence of what can be achieved in PA promotion within primary care settings. Nonetheless, several challenges exist. First, there is no clear agreement among PC organizations and evidence-gathering agencies on the recommendations for PA promotion within the PC context. For example, the United States Preventive Services Task Force (USPSTF) currently recommends selective PA counseling rather than incorporating the message into routine practice in the general population (Moyer and U.S. Preventive Services Task Force, 2012); the United Kingdom's National Institute for Health and Clinical Excellence (NICE) also recommends that the National Health Service provides brief advice to adults who have been assessed as being inactive (National Institute for Health and Care Excellence, 2013), but the Royal Australian College

http://dx.doi.org/10.1016/j.ypmed.2014.09.012 0091-7435/© 2014 Elsevier Inc. All rights reserved.

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

2

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

of General Practice (Royal Australian College of General Practice, 2012) proposes that all adults and children should receive advice (Table 1). Second, the specific content and delivery format and the most effective elements of PC interventions promoting PA remain unclear (Hillsdon, 2013). And third, clarification is needed about supporting evidence to guide GPs in prioritizing their behavioral counseling efforts. Certain patients may be predisposed to benefit from these efforts based on, for example, their risk factor profile or readiness for change (as suggested in USPSTF guidelines) or their inactive lifestyle habits (as stated in the NICE recommendations). The aim of this descriptive review is to summarize the evidence of the effectiveness of PA promotion interventions in the PC setting designed to increase PA levels of adult patients, as presented in systematic reviews and meta-analyses published from 2002 to 2012. Further, it attempts to determine the intervention components or strategies that have proven to be the most effective and the patient characteristics that could guide PC professionals to prioritize their efforts or maximize the impact of interventions. Finally, the implications of incorporating the available evidence into practice will be discussed. Methods The present review has been registered in PROSPERO: CRD42013004413. Review questions 1) What is the current state of evidence regarding the effectiveness of PApromoting interventions in the PC setting to increase the PA level of adult patients? 2) Which intervention components or strategies have been shown to be most effective? 3) Which sample characteristics are associated with higher effects? Data sources and search strategy A search was performed in MEDLINE, EMBASE, and Cochrane Library databases to identify systematic reviews and meta-analyses published from 2002 to 2012 that assessed the effectiveness of PA-promoting interventions in the PC setting. A search strategy was developed using free text and subject

heading terms related to the behavior or habit studied (physical activity, exercise, leisure or motor activities), to interventions that offered counseling or assistance (counseling, patient education, behavior change, health promotion), and to the specific study design (meta-analysis or systematic review) that was the object of the search. In order to maximize search sensitivity, we did not initially filter reviews by terms related to “primary care.” Instead, we thoroughly applied the inclusion/exclusion criteria to each study referenced. The search was not limited to any language or country of origin. In addition to the databases searched, the reference lists for all of the selected reviews/meta-analyses were consulted to identify potentially eligible studies. Furthermore, a rapid Web search (“physical activity” and “primary care”) was performed. The year 2002 was selected as the starting point in order to focus on newer reviews that may better represent current working conditions, trends, and procedures in PC. Inclusion criteria 1) Participants/population: adults aged 18 years and older; 2) Intervention: any intervention performed or initiated in a PC setting with the goal of increasing the PA level or participation of sedentary or insufficiently active adults; 3) Comparison group: no intervention control, usual care control, or alternative intervention control; and 4) Context: interventions initiated in a PC context with PC professionals as main intervention agents. The present study follows the definition of PC as “level of a health service system that provides entry into the system for all new needs and problems, provides person-focused (not disease-oriented) care over time, provides care for all but very uncommon or unusual conditions, and coordinates or integrates care provided elsewhere by others” (Starfield, 1998); 5) Types of study: literature reviews, systematic reviews, meta-analyses; and 6) Primary outcome: increase in PA level or proportion of patients meeting predefined PA level, with at least one post-intervention follow-up measurement. Exclusion criteria 1) Clinical practice guidelines or recommendations involving no literature search and review of studies analyzing evidence; 2) Reviews in which primary studies carried out in PC did not constitute at least 50% of the included articles; 3) Studies conducted in settings that were not generalizable to primary care, including inpatient care, emergency departments, or occupational settings; 4) Reviews of secondary or tertiary prevention, or population studies focused only on pathology, as the presence of chronic illness may cause patients to be more

Table 1 Recommendations for physical activity promotion within the Primary Health Care context from organizations and evidence-gathering agencies. Royal Australian College of General Practitioners (Royal Australian College of General Practice, 2012)

All adults should be advised to participate in 30 min of moderate activity on most, preferably all, days of the week

US Preventive Services Task Force (Moyer and U.S. Preventive Services Task Force, 2012)

Existing evidence indicates that the health benefit of initiating behavioral counseling in the primary care setting to promote physical activity is small. Clinicians may choose to selectively counsel patients rather than incorporate counseling into the care of all adults in the general population

NICE (National Institute for Health and Care Excellence, 2013)

Advise adults who have been assessed as being inactive to do more physical activity

Interventions that have shown short-term benefit in changing physical activity include: a) patient screening to identify current level of activity (including use of a pedometer) and readiness to be more active b) provision of brief advice or counseling on exercise c) supporting written materials and/or written prescription for exercise d) pedometer step target of 10 000 steps per day, or 2000 more than at baseline Studies of medium- and high-intensity behavioral counseling interventions have shown beneficial effects on behavioral and intermediate health outcomes. Medium-intensity interventions involved a range of 3 to 24 phone sessions or 1 to 8 in-person sessions. High-intensity interventions involved a range of 4 to 20 in-person group sessions and were the only interventions to report sustained benefits beyond 12 months. No high-intensity interventions and few medium-intensity interventions involved primary care clinicians as the providers of the intervention Tailored advice to: a) motivations and goals; b) current level of activity and ability; c) circumstances, preferences and barriers to being physically active; d) health status Provide information about local opportunities to be physically active for people with a range of abilities, preferences and needs. Consider giving a written outline of the advice and goals that have been discussed. Follow up when there is another appointment or opportunity.

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

3

Fig. 1. PRISMA search algorithm.

responsive to interventions and care may imply the participation of further specialists and additional resources not generally accessible within primary care context (Conn et al., 2011; Hudon et al., 2008); and 5) Exercise referral schemes, as intervention is mainly performed by a third-party service provider and may include supervised exercise training.

Review selection Two team members reviewed the abstracts and summaries, applied the established inclusion and exclusion criteria, and independently made an initial selection of studies identified by the search strategy. Studies not meeting the exclusion criteria were further reviewed by accessing the full text. Potential inconsistencies were identified, discussed, and resolved by consensus.

Study quality assessment and grading of evidence Two team members independently assessed the quality of each selected study and the level of evidence for the conclusions reached, using the nineitem OQAQ scale (Oxman and Guyatt, 1991) designed to assess the methodological quality of systematic reviews and the Scottish Intercollegiate Guidelines Network (SIGN) classification system for degree of evidence (Scottish Intercollegiate Guidelines Network (SIGN), 2001). The SIGN system assesses the risk of bias associated with a particular piece of evidence in a hierarchy ranging from grade 1 (evidence from meta-analysis and/or RCTs) to grade 4 (evidence based on expert opinion). Indicators were assigned to indicate methodological quality (++ high quality; + acceptable quality, or − low quality). Any disagreement was resolved by discussion and consensus.

Data extraction The key information was extracted using two forms. Descriptive information and methodological characteristics (objectives, design, inclusion/ exclusion criteria, and description of the intervention and the comparison group, outcome, period covered) were recorded on one form and the second form focused on results, quality of work, and level of evidence.

Analysis and data reporting The aim was to summarize and state the evidence provided by the selected reviews and meta-analyses. Therefore, a narrative format was selected for reporting and a systematic report structure was followed to the extent possible. In accordance with reporting guidelines for systematic reviews, a PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist for the present review is detailed in the Appendix 1. No statistical analysis was conducted.

Results The search strategy identified 1664 records. Six additional studies were identified from the references in the selected publications and three studies were located by the Web search. After reviewing these titles and abstracts, 91 publications were considered potentially eligible for inclusion. After a full-text review of these potentially eligible articles, only 10 studies met all of the inclusion criteria. The main reasons for excluding articles from the present review were the absence of a literature search with a review and analysis of evidence, a study population with some specific disease, fewer than 50% of the primary studies carried out in PC settings, inadequate outcomes measures, or study objectives that were irrelevant to the aim of the present study (Fig. 1). Review characteristics Of the 10 studies included for analysis (Table 2), five are systematic reviews with a meta-analysis of aggregated effects (Conn et al., 2011; Campbell et al., 2012; Orrow et al., 2012; Lin et al., 2010; Hillsdon et al., 2005), three are systematic reviews (van Sluijs et al., 2004; Smith, 2002; Eden et al., 2002), and two are reviews of the literature (Sørensen et al., 2006; Tulloch et al., 2006). Only two of these reviews included exclusively randomized clinical trials (Hillsdon et al., 2005; Eden et al., 2002). In the majority of the reviews, the comparison group is usual care or no intervention (Campbell et al., 2012; Orrow

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

4

Study

Type of Review

Objective (O) and Questions (Q)

Inclusion/Exclusion Criteria

Outcome Measures

Review Period

1. Campbell F et al., 2012

Systematic review with meta-analysis

O1: to evaluate the effectiveness of brief counseling interventions in PHC to promote PA in adults O2: to evaluate which aspects of the intervention contribute to effectiveness

Effect size (OR or standardized mean difference) of long-term change in the level of PA or cardiorespiratory capacity

1990-2012

2. Orrow et al., 2012

Systematic review with meta-analysis

O1: to determine the sustained effectiveness of PHC interventions promoting PA in sedentary patients O2: to determine the effectiveness of these PHC interventions

Effect size (OR or standardized mean difference) of long-term change in the level of PA or cardiorespiratory capacity (N = 12 months)

1999-2010

3. Conn et al., 2011

Systematic review and meta-analysis

Effect size of change in the level of PA

1960-2007

4. Lin et al., 2010

Systematic review and meta-analysis

O: To determine the effect of interventions designed to increase PA in healthy adults Q1: What is the overall effect of interventions designed to increase PA after completion of the intervention? Q2: Do the effects of interventions vary as a function of the intervention, methodology or sample characteristics? O: To evaluate the effectiveness of interventions promoting PA and diet in the prevention of cardiovascular diseases

Design: Randomized or nonrandomized clinical trials, in which the intervention was carried out by PHC professionals in not more than 30 min. The control group was defined as usual care, and varied between studies. Participants: Adult patients older than 19 years Design: Randomized clinical trials with at least 12-month follow-up of interventions to increase the level of PA, compared to a comparison intervention or no intervention Participants: Sedentary patients older than 16 years recruited in PHC settings. Design: Studies of interventions to increase PA. Participants: Healthy adults

Effect size (RR or standardized mean difference) of change in the level of PA

2001-2008

5. Sorensen et al., 2006

Review of the literature

Proportion of patients completing 150 min of PA per week. Increase in maximal oxygen consumption.

1980-2005

O1: To evaluate the effect of prescribing PA on increasing PA levels or physical fitness and whether more intensive interventions obtain better outcomes Q2: Is the prescription of PA feasible in PHC and accepted by patients? Q3: Is the prescription of PA cost-effective?

Design: Randomized or nonrandomized clinical trials with a comparison group. Intervention with counseling or behaviorbased assistance carried out or feasible for implementation in PHC to promote improved PA/diet. Group comparison: no intervention, usual care or minimal intervention. Studies with 6 months or less of follow-up were excluded. Participants: unselected adults. Trials were excluded if N50% of the sample had hypertension, hyperlipidemia, diabetes or cardiovascular disease. Design: Studies of the prescription of PA, defined as a more intensive intervention than simply counseling (eg, additional counseling, written materials, telephone follow-up, supervised activity) carried out by a GP or other PHC professional, with outcome measures for PA or fitness. Interventions with institutionalized patients or b6 months follow-up were excluded Participants: Sedentary adult patients with “signs of lifestyle-related diseases”

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

Table 2 Characteristics of Review Articles Analyzed.

Review of the literature

O: To examine differences in the effectiveness of PA interventions by family doctor, allied health professional, or in combination

7. Foster et al., 2005

Systematic review and meta-analysis

8. Van Sluijs et al., 2004

Systematic review

O: To evaluate the effectiveness of interventions promoting PA in adults in comparison with no intervention, minimum intervention, or alternative intervention (controls) Secondary aims: Q1: Are more intensive interventions more effective? Q2: Are some components associated with better outcomes? Q3: Are the effects short-term on long-term? Q4: Do certain types of activity obtain better outcomes? Q5: Are interventions more successful with a certain type of participants? O: To evaluate the effectiveness of PHC interventions promoting PA that are based on the Trans-theoretical Model of Behavioral Change

9. Smith et al., 2002

Systematic review

O1: To determine whether PHC interventions promoting PA (or multi-risk) are effective to increase participation in PA O2: To examine differences in outcome based intensity, setting, and patient characteristics

10. Eden et al., 2002

Systematic review

O1: To determine whether counseling adult patients in PHC increases and maintains PA levels O2: To determine which types of interventions are most effective

Design: Randomized clinical trials or quasiexperimental studies promoting PA and with an outcome measure for PA. Interventions carried out by a GP, a GP in combination with an allied health professional, or only an allied health professional. Comparison group: not specified. Participants: adults recruited from primary care Design: Randomized clinical trials comparing different interventions promoting PA, with at least 6 months of follow-up, analyzed by intention to treat or with b20% losses to follow-up. Intervention: carried out by one or more health professionals (GP, nurse, health educator, therapist, etc.) Comparison group: no intervention, minimal intervention, or alternative intervention. Participants: adults with no limiting conditions

Design: Trials promoting PA initiated in PHC, based on the stages of change model, with at least one follow-up measure. Intervention: Verbal or computer-based counseling or printed materials. Comparison group: no intervention or usual care. Participants: adults recruited in PHC Design: Randomized clinical trials or quasiexperimental controlled studies promoting PA or multiple healthy habits and with a measurement of PA outcomes. Intervention: Strategies involving counseling and support, in-person or other follow-up, written prescription, and other behavioral techniques or printed materials. Comparison group: no intervention or usual care. Participants: adults recruited in PHC Design: Controlled clinical trials with PA outcome measures. Intervention: at least one component of the intervention must have been carried out by PHC professionals. Comparison group: no intervention or usual care. Participants: adults in primary care

Change in the level of PA

2000-2005 (including studies of previous reviews)

Effect size (RR or standardized mean difference) of change in the level of PA or fitness

Until 2004

Positive outcome (significant difference) in favor of the intervention, at short (b6 months), mid (6 months) and long term (N6 months)

Until 2002

Positive outcome (significant difference) in favor of the intervention

1966-2002

Proportion completing the recommendations or long-term changes in the level of PA (N6 months after randomization)

1994-2002

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

6. Tulloch et al., 2006

5

6

Study

Comparisons

Method of comparison

Number of studies

Outcome Variable

Length of Follow-up

Campbell et al., 2012

Brief interventions promoting PA compared to: 1) usual care 2) more intensive interventions

Aggregate effect based on metaanalysis of randomized clinical trials

1) 16 2) 5

Effect size (OR o standardized mean difference) of change in the level of PA or cardiorespiratory capacity

Orrow et al., 2012

Aggregate effect based on meta1) Interventions promoting PA vs comparison intervention group or analysis of randomized clinical trials no intervention 2) referral intervention vs comparison intervention group or no intervention 3) Analysis of subgroups: intervention effect vs. comparison intervention or no intervention

Total 15

Effect size (OR or standardized mean difference) in long-term change in the level of PA or cardiorespiratory capacity

Interventions promoting PA compared to a control group

206

From 4 weeks to 1) Seven of 16 studies reported significant 12 months differences. The 8 studies with quantitative measurements had an estimated standardized mean difference of 0.17, 0.06-0.28, I2 = 69%, in favor of the intervention. The 4 studies with dichotomous measures showed a positive effect in RR: 1.30, 1.12-1.50; I2 = 66%. 2) Of the 5 studies included, 3 found that intensive interventions were more effective. Nonetheless, the aggregate effect was not significant and no additional benefit was demonstrated. Other evidences reported: The most-used strategies were support in shaping intention, information about consequences, use of follow-up techniques and establishment of objectives. Recommendations include intervention with sedentary individuals; adding written materials do not improve the effect; no differences observed by sex or age group; lesser effect was observed in disadvantaged populations. Health professionals are more likely to intervene with motivated patients and when they think PA is important, they are active or have more confidence and training. Lack of time was the major barrier. N = 12 months 1) Small to moderate positive effect derived from 13 trials (OR: 1.42, 95%CI 1.17-1.73, I2 = 43%; standardized mean difference: 0.25, 95%CI 0.110.38, I2 = 70%, NNT 12 (95%CI 7–33). No significant cardiorespiratory effect. Interventions included printed materials and verbal counseling by health care professionals, in person or by telephone on multiple occasions. 2) No significant effect 3) Six studies comparing intervention vs no intervention showed a moderate positive effect (OR: 1.72, 95%CI 1.39-2.18, I2 = 21%; Standardized mean difference: 0.36, 95%CI 0.280.43). Seven studies of intervention vs comparison showed no significant effect (OR:1.18, 95%CI 0.85-1.48, I2 = 9%; Standardized mean difference: −0.03, 95%CI −0.17-0.11, I2 = 0%) Note: The main limitation is that most studies used questionnaires to measure outcomes 1) Mean effect size of 0.19 in comparisons of Postintervention vs control, post-intervention intervention (I2 = 0.67) and pre-post intervention vs control (not specified) (I2 = 0.59) 2) Various effect moderators are associated with larger effect sizes: behavioral interventions compared to cognitive interventions or those based on social cognitive theory or trans-

Conn et al., 2011

Effect size (OR or standardized mean difference) of change in the level of PA based on meta-analysis of intervention studies

1) 13 2) 3

Effect size of change in the level of PA

Results

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

Table 3 Summary of Evidence: Effectiveness of Interventions.

1) Low-intensity interventions promoting PA (total contact b30 min) compared to control group 2) Moderate-intensity interventions promoting PA (30 min to 6 contact hours) compared to control group 3) Interventions promoting PA and diet Note: Most of the moderate- and high-intensity interventions were carried out by specialized professionals, not by PHC professionals

Aggregate effect based on metaanalysis of randomized clinical trials

1) 11 2) 19 3) 17

Effect size of change in the level of PA

Sorensen et al., 2006

Intervention of a PA prescription beyond basic counseling, compared to a low-intensity intervention or no intervention

Descriptive summary of outcomes and aggregate effect (no information about how they were calculated)

12 studies with no data on their design

6-12 months Proportion of patients who complete 150 min of PA per week. Increased maximal oxygen consumption

Tulloch et al., 2006

1) 7/19 1) Intervention carried out by general Descriptive summary of outcomes (no information about how they were 2) 5/19 practitioners (GPs) compared to 3) 7/19 calculated) control group 2) Intervention carried out by GP in combination with an allied health professional compared to control group 3) Intervention carried out by allied health professional compared to control group

Percentage of studies with positive outcomes in PA changes compared to control group

Short-term (b6 months) or long-term (N6 months

7

(continued on next page)

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

Up to 12 months

Lin et al., 2010

theoretical models. (Highlights: establishment of objectives, self-monitoring, feedback, modeling, indications and consequences, prescriptions); individualized interventions carried out face-toface. Greater effect in chronic patients or those diagnosed with a disease.) 1) Four of 11 studies obtained an effect in some PA variable. The effect size of overall reported PA level was null (0.08, −0.01-0.18, I2 = 40.7%; k = 8; N = 6,288). The only high quality study obtained a difference of 17.5 min/week of moderate to vigorous PA. Of the 3 studies reporting over-compliance with recommended PA at 6 and 12 months, the overall effect (RR) was 1.25 (1.11-1.41, I2 = 0%, N = 4,289), a 4% difference in absolute risk and NNT of 25. 2) Overall effect on PA level was 0.19 (0.12-0.27, I2 = 49.9%; k = 17; N = 6,808). The 9 studies that reported minutes per week had a difference of 38 min/wk compared to controls. Effect on adherence to recommendations 1.22 (1.07-1.40, I2 = 59%; k = 6; N = 4,183), 7% difference in absolute risk and NNT of 14–15. 3) The only low-intensity study obtained no effect. The global effect of the level of PA in studies of moderate and low intensity interventions is significant (0.20, 0.08-0.33, I2 = 45.7%; k = 8; N = 2,189; 0.26, 0.14-0.37, I2 = 0%; k = 4; N = 1,152) In half of these studies, the level of PA Increased significantly in patients who received a PA prescription. The prescription had a moderately positive effect in 10% of patients, and the average VO2max value increased 5%-10% compared with control groups after 6 to 12 months. GP participants tend to be volunteers with a positive attitude. Few studies included a random sample of GPs, making it difficult to generalize the results to all GPs. Patient acceptance was very good. Of the short-term studies, 3 of 6 (50%) involving only GPs obtained positive outcomes, as did 2/3 (67%) of the combined interventions and both (100%) of those carried out only by allied health professionals. The success of the last two types of study was due to a more intensive intervention and a higher degree of specialized professional training. In the long term, positive outcomes were obtained in 2/4 (50%) of the GP interventions, 2/3 (67%) of the combined interventions, and 5/7 of those carried out by allied health professionals. The successful GP interventions were longer and had multiple follow-up contacts. In the remaining studies, better outcomes were associated with multiple components (written prescriptions, specific objectives, follow-up sessions, community resources) Conclusion: It is less important who carries out

8

Study

Comparisons

Method of comparison

Outcome Variable

Length of Follow-up

29 RCT

Effect size of change in the level of PA or physical fitness

N6 months

Descriptive summary of outcomes

13 studies, 10 of high quality (10 RCTs, 8 of high quality)

Significant differences between comparison groups

Short-term (b6 months), mid-term (6 months) and long-term (N6 months)

Descriptive summary of outcomes

1) 8 (6 RCTs) 2) 12 (8 RCTs)

Significant differences between comparison groups

Short-term (b6 months), mid-term

Foster et al., 2005

Interventions promoting PA in adults, Aggregate effect based on a metaanalysis of randomized clinical trials in comparison with no intervention, minimum intervention or alternative intervention (controls)

Van Sluijs et al., 2004

Intervention promoting PA based on the stages of change model, compared to no intervention or usual care

Smith et al. 2002 1) Intervention promoting multiple healthy habits compared to control group

Number of studies

Results the intervention than its intensity and the specific qualifications of the health professional 1) Seven of 19 studies with continuous assessment obtained positive outcomes. The aggregate effect was moderate, with significant heterogeneity (0.28, 0.15-0.41, I2 = 83.5%). Considering only the high-quality trials, the effect was positive and without heterogeneity (0.11, 0.04-0.17). Only 2 of 10 studies obtained positive outcomes in adherence to a predetermined level of PA, with significant heterogeneity. The aggregate effect (OR) was 1.33 (1.03-1.72, n = 53.4%). Considering only the high-quality trials, the effect was positive and without heterogeneity (1.48, 1.07-2.06). 2) Five of 11 studies obtained positive outcomes in physical fitness (0.52, 0.14-0.90). Conclusion: There is evidence of a moderate effect at short- and mid-term of interventions promoting PA. The interventions with professional assistance in starting an exercise program and ongoing support are the most effective. Three of the 9 high-quality studies included obtained significant differences in the short term, in favor of the intervention. Only 1 (a high-quality RCT) of a total of 5 studies showed significant differences in the mid- term. Only 1 (a high-quality RCT) of a total of 5 studies showed significant differences in the long term. Conclusion: There is no evidence of that a PHC intervention promoting PA that is based on the stages of change model is effective in the short, mid or long term. 1) Four of the 7 studies with long-term follow-up showed significant differences and one of them in the short term as well.

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

Table 3 (continued)

Eden et al., 2002

Interventions promoting PA in which at least one component is carried out by a PHC professional, compared to usual care or an alternative intervention

(6 months) and long-term (N6 months)

Descriptive summary of outcomes

7 RCTs and 1 NRCT

Proportion of adherence Long-term (N6 months) to recommendations or changes in the level of PA in the long term (N6 months post randomization)

Intervention intensity ranged from one evaluation session with counseling to multiple sessions with telephone follow-up. In three studies, the intervention was carried out by nurses or GPs. Only one intervention was part of routine PHC practice. 2) Six of 8 studies (3 of 4 RCTs) obtained significant differences in the short term. Three of 6 (2 of 4 RCTs) found significant differences in the mid-term, and 2 of 5 in the long term, although one RCT reported a difference only for women. Three of 4 studies with outcomes in the short term found no differences, or observed a decreased outcome at mid and long term. Adding extra components to the intervention (eg, counseling + prescription) produced better short-term results. Better outcomes were obtained in sedentary patients. Only 3 studies were carried out in routine PHC practice. Of 6 trials with a group receiving usual care, the only high quality study found no significant differences. Of the 5 studies with acceptable quality, 3 obtained a significant increase in PA. Components of the interventions included: counseling, help with perceived efficacy and barriers, educational materials, referral to community resources and written prescriptions In both trials that compared different interventions, it stands out that written prescriptions obtained better outcomes than counseling/advice and that women required more intensive intervention. Conclusion: There is insufficient evidence to conclude that simply offering PA advice in PHC is effective.

Note: PHC: Primary Health Care; PA: Physical activity; OR: Odds Ratio; RR: Risk Ratio; RCT: Randomized Control Trial; GP: General Practitioner; I2: Heterogeneity; vs: Versus; CI: Confidence Interval; NNT: Number needed to treat; k: Number of studies included; N: Number of patients included; min/wk: Minutes per week; VO2max: Maximum Oxygen Volume; NRCT: Non Randomized Control Trial

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

2) Intervention promoting PA habits compared to control group

9

10

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

et al., 2012; van Sluijs et al., 2004; Smith, 2002; Eden et al., 2002) and two reviews (Lin et al., 2010; Hillsdon et al., 2005) also included the use of an alternative intervention or minimum intervention as a control. The three remaining reviews did not specify the type or content of the control (Conn et al., 2011; Sørensen et al., 2006; Tulloch et al., 2006). Seven of the reviews attempted to determine which of the intervention or patient characteristics were associated with better outcomes (Conn et al., 2011; Campbell et al., 2012; Lin et al., 2010; Hillsdon et al., 2005; Smith, 2002; Eden et al., 2002; Tulloch et al., 2006). Three studies included multi-risk interventions (Lin et al., 2010; van Sluijs et al., 2004; Smith, 2002), and the rest evaluated only interventions designed for PA promotion. Study quality Appendix 2 presents the results of the evaluation of the methodological quality of the studies included. The five reviews that included a meta-analysis (Conn et al., 2011; Campbell et al., 2012; Orrow et al., 2012; Lin et al., 2010; Hillsdon et al., 2005) obtained a quality score greater than 16 points on the OQAQ and all contributed high-quality causal evidence (1++) in their conclusions. Of the remaining studies, 3 systematic reviews (van Sluijs et al., 2004; Smith, 2002; Eden et al., 2002) and one of the literature reviews (Tulloch et al., 2006) contributed causal evidence of average quality (1+), with OQAQ scores ranging from 15 to 17. Only one literature review had a low grade of evidence (Sørensen et al., 2006). Evidence synthesis Effectiveness of PA promotion interventions in PC High quality studies. High-quality causal evidence of a positive effect was shown in five systematic reviews with an aggregated effect analysis (Conn et al., 2011; Campbell et al., 2012; Orrow et al., 2012; Lin et al., 2010; Hillsdon et al., 2005) (Table 3). In general, the trend is to report an overall positive effect, which is greater when the comparison is with a control group or usual care. Four of the five obtained a small to moderate mean standardized effect (0.17−0.28). These positive results must be interpreted with caution due to the high degree of heterogeneity (I2 range: 67% to 83.5%), with a moderate effect (range of RR: 1.22−1.42) on achieving the predetermined PA level. In the sensitivity analysis, considering only high-quality trials (Hillsdon et al., 2005) or the comparison of the intervention to no intervention (Orrow et al., 2012), the positive effect in favor of PA promotion interventions in PC was maintained or increased, with a low degree of heterogeneity. The effect analysis of the intensity of the intervention by subgroups seems to show that more intensive interventions (longer than 30 min, mainly carried out by specialized professionals) have a greater impact on the overall aggregated effect on PA level, but not on adherence to the predetermined PA level (Lin et al., 2010). Only one high-quality meta-analysis (Orrow et al., 2012) evaluates and reports a positive effect of interventions to promote PA in the PC setting beyond 12 months. Only two of the studies (Orrow et al., 2012; Lin et al., 2010) provided estimates of the number of insufficiently active patients needed to treat (NNT) with a PA promotion intervention to achieve one patient completing the recommended levels. The reported NNT ranged from 12 to 25. Medium or low quality studies. Of the three systematic reviews of average quality and with a moderate or low degree of evidence (van Sluijs et al., 2004; Smith, 2002; Eden et al., 2002), only one (Smith, 2002) reported acceptable evidence that the interventions that addressed PA behaviors in PC patients could achieve improvements, especially when the patients were insufficiently active. Nonetheless, the generalizability of the results is limited because most of the studies were selected samples. In another of these systematic reviews (van Sluijs et al., 2004), the results were mixed; in addition, several studies had methodological limitations. This leads to the conclusion that the

available evidence of the effectiveness of counseling to promote PA in the PC setting is inconclusive. The third review, despite finding results in one of the studies, concluded that there was no evidence of any short-term, mid-term, or long-term effect of PA promotion interventions based on the stages of change model and carried out in PC settings (Eden et al., 2002). Notably, these reviews were the oldest of those included and analyzed studies published before 2002. With respect to the remaining two literature reviews (Sørensen et al., 2006; Tulloch et al., 2006), the average-quality study (Tulloch et al., 2006) found positive results for PC interventions, both in the short and medium term. The authors conclude that interventions combining health professionals and allied health professionals with specialized training obtain the better results, reducing the demand on the family doctor's time and delivering more intensive specialized care. The other review of literature (Sørensen et al., 2006), which had a low quality score, concluded that the prescription of exercise – defined as an intervention that goes beyond simple advice from the PC physician or other health professionals – increases the level of PA and of physical fitness in sedentary patients with signs of lifestyle-related diseases. Subgroup analyses regarding specific characteristics of interventions or patients Four high-quality (++) studies conducted several subgroup analyses regarding the association between intervention or patient characteristics and effectiveness (Conn et al., 2011; Campbell et al., 2012; Lin et al., 2010; Hillsdon et al., 2005). Campbell et al. (Campbell et al., 2012) reported no clear relationships between PA outcomes and specific features of the intervention content, setting, and delivery. In particular, little evidence of additional benefit from a more extensive intervention was found. Specifically, they found inconclusive evidence when comparing brief and very brief interventions. Although more intensive interventions seemed to have better PA outcomes, there was no clear benefit from the addition of further interventions (e.g., written material) to support brief advice. Regarding patient characteristics, the authors found insufficient evidence to allow conclusions about the impact of sex and age group on intervention effectiveness. Brief advice may be less effective in disadvantaged populations. Although there was insufficient evidence to generate a conclusion, the authors recommend targeting advice to the sedentary population. With respect to the analysis of behavior change techniques, the authors conclude that the most-used techniques were prompt intention formation and goal setting, providing general information on behavior-health links and information on consequences of limited PA, and use of follow-up prompts. No association with effectiveness was reported. No clear evidence regarding the role of behavior change models in designing effective interventions was found. In the study by Conn et al. 2011), various moderator variables were associated with greater effects on PA change. Specifically, behavioral interventions had better outcomes compared to cognitive interventions. Behavioral strategies include goal setting, self-monitoring, feedback, consequences, exercise prescription, and cues to action. Larger PA effects seemed to be associated with interventions delivered directly to individuals (face-to-face vs mass-media interventions and community-wide interventions), interventions delivered by project staff (vs train-thetrainer models), PA behavior modeled by research staff, standardized interventions (vs individually tailored interventions), and interventions not based on Social Cognitive Theory or the Trans-theoretical model. Lately, higher results were found in chronically ill patients as compared to healthy population. However, as the authors comment, all previous conclusions should be considered as exploratory due to the limitations within the moderator analyses performed. Analyses from the study of Lin et al. (2010) were stratified by intervention intensity: low-intensity trials (30 min or less of contact with providers); medium-intensity trials (more than 30 min but less than 6 h of contact); and high-intensity trials (estimated to involve more than 6 h of contact). Although there was significant statistical

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

heterogeneity across interventions, the authors concluded that there appeared to be an increasing effect size with intervention intensity. Foster et al. (Hillsdon et al., 2005) examined a potential modification of effects depending on particular participant groups or the following five categories of interventions: the nature of the first patient-intervention contact, degree of program supervision, frequency of intervention occasions, frequency of follow-up contacts, and type of follow-up contacts. Due to the clinical and statistical heterogeneity of the studies, only limited conclusions can be drawn about these secondary analyses. No clear effect of specific patient characteristics could be extracted. Regarding intervention characteristics, results seemed to be more favorable when PA was self-directed with some professional guidance and when there was on-going professional support, reducing the heterogeneity of effects. Conclusions from two low-quality systematic reviews (Smith, 2002; Eden et al., 2002) and from one low-quality literature review (Tulloch et al., 2006) also point toward a higher effect size with more intensive interventions, although another low-quality literature review (Sørensen et al., 2006) found little evidence to support this finding. However, these reviews note that very few of the analyzed studies followed routine PC conditions, in some cases involving additional resources or specifically qualified allied health professionals and thereby limiting the generalizability of results. In the case of Tulloch et al. (2006), they concluded that the effect depends not on who performed the intervention, but rather on the intensity of the intervention and the qualification and training of the intervention agent. Successful interventions performed by family physicians incorporated more time and multiple follow-up contacts. In combined professional interventions and in those performed only by allied health professionals, better outcomes are associated with multiple intervention components, such as written prescriptions, goal setting, and follow-up contacts. Only two studies analyzed the role of the theoretical models applied (Conn et al., 2011; van Sluijs et al., 2004), identifying no favorable effect from interventions based on the trans-theoretical model of stages of change and the social-cognitive model (Campbell et al., 2012).

11

the estimated NNT for smoking cessation advice by physicians of 50 to 120 as a reference (Stead et al., 2013), NNT estimates of 12 to 25 for PA promotion interventions in PC seems to be highly relevant. Therefore, a clear recommendation of PA promotion in PC interventions to increase the PA level of insufficiently active patients can be made. In contrast, the state of evidence regarding the association between intervention components, patient characteristics, and effectiveness of interventions precludes any clear statement on these two issues. Subgroup analyses by some of the included reviews suggest that the most often-used and effective intervention strategies are behavioral in nature, such as providing a prescription, establishing objectives, providing feedback, encouraging self-monitoring, and ongoing follow-up and professional support. More intensive interventions seem to deliver better outcomes. Nonetheless, they tend to involve additional professional or financial resources and are therefore less generalizable to routine PC conditions. Although the present study showed that provider-based PA interventions have a small but positive effect on changing the PA levels of PC patients, training physicians in the use of behavior change strategies and coordination with clinical or community resources could be a possible way to maximize their impact. The pooled review of reviews and meta-analysis of PA promotion in PC that evaluated the association between patient characteristics and effectiveness of interventions offers no clear evidence (Conn et al., 2011; Campbell et al., 2012; Lin et al., 2010). Some patient characteristics were highlighted in recommendations made to adapt and target PA interventions, such as risk factors, readiness for change, or PA level (Lin et al., 2010; National Institute for Health and Care Excellence, 2013). Although there is insufficient evidence to support it, a few of the reviewed studies recommend targeting advice to insufficiently active or sedentary population as an approach that may be more successful (Campbell et al., 2012). This will imply the development of a feasible strategy and/or procedures for assessing PA to be implemented under routine PC conditions. Further evidence is warranted regarding personal characteristics associated with higher effectiveness, in order to establish a basis for targeting strategies.

Discussion The present study of a decade of systematic reviews and metaanalyses found that interventions designed for PA promotion in the PC setting that are based on advice from health professionals have a small to moderate positive effect on increasing PA levels. Interventions that include multiple behavioral change strategies (establishing objectives, providing feedback, writing PA prescriptions, etc.) seem to obtain better outcomes. A certain level of evidence showed a long-term effect, although the effects tend to disappear without follow-up or continued support. Nonetheless, with respect to the intensity of counseling interventions and the variability of effects depending on the characteristics of the sample (sex, age, socioeconomic level), no clear associations were observed. The interventions with referral schemes that were initiated in the PC context seemed to have a small positive effect on increasing PA levels, especially when compared to usual care. These results must be interpreted with caution because the evidence is based primarily on self-reported measures of PA levels and selected populations that are not necessarily representative of PC populations. Finally, very few studies were carried out under routine PC working conditions. The present review summarized and graded the evidence of PA promotion in the context of PC published over a recent decade. The majority of the systematic reviews and meta-analyses that were graded as high quality showed a clear positive effect in favor of counseling interventions to increase PA levels and to meet the recommended PA levels (standardized mean effect range: 0.17–0.28; RR/OR range: 1.22–1.42). Despite significant heterogeneity in the reported effects for the main comparisons, further sensitivity analysis with higher quality studies showed maintenance and even improvement of effects while reducing heterogeneity. Regarding clinical relevance of effects, and considering

Limitations The present study reviewed the evidence for PA promotion interventions in the PC context as derived from existing reviews and metaanalyses published from 2002 to 2012. Consequently, it is possible that evidence provided by single studies was missed. It also must be noted that reviews in which less than half of the reviewed studies were from the PC context were excluded; therefore, some potentially interesting insights might have been lost. Overlap among reviews and heterogeneity of studies hampers the possibility of determining an effect size for PA promotion in PC interventions. Another limitation noted by authors of the included reviews and meta-analyses is that the majority of reported evidence relies on self-reported measures of PA, in some cases with selected samples not representing the PC population. Special caution should be used in interpreting the subgroup analyses, considering them as providing only associative evidence. Finally, it must be noted that most studies evaluating PA interventions in PC and included in evidence reviews have been conducted in the United Kingdom, North America, Northern Europe, Australia, and New Zealand. Therefore, the available evidence may have limited generalizability to other countries. Conclusions Supported by clear evidence of small but positive results provided by several high-quality reviews, we advocate for interventions in PC settings designed to increase PA levels of patients. Interventions that include multiple behavioral change techniques and those targeted to insufficiently active or sedentary patients seem to have better results.

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012

12

A. Sanchez et al. / Preventive Medicine xxx (2014) xxx–xxx

However, some important questions still could not be answered due to lack of specific information about reported trials or because pooled analyses did not provide sufficiently clear evidence. First, there are important gaps in knowledge related to specific characteristics of effective interventions; for example, the most effective intervention effective format and content and the recommended duration of the PA-promoting intervention, including follow-up, remains unclear. Furthermore, the association between patient characteristics and effective interventions is not sufficiently well established to determine the most efficient intervention strategy for individual patients (e.g., opportunistic, selective, population-based), especially when trying to conduct PA interventions in routine clinical settings. In fact, routine clinical implementation of PA interventions is the main gap to be addressed, as the majority of evaluated interventions ceased when trials ended. This is due to the difficulties of integrating them within routine PC delivery services. Furthermore, the available evidence has been provided from selected samples, with only a few of the studies conducted under routine PC conditions and no information provided about external validity. This limits the generalizability of intervention outcomes to real-world settings. As the available evidence of the efficacy of PA interventions in the primary care setting appears to be growing, future research should focus on implementation trials that occur under routine primary care conditions, in order to provide knowledge about factors that affect external validity, such as determinants of practice and context-related information. Implementation trials – that is, those aimed at designing specific strategies to enhance the population reached and encourage the adoption and implementation of evidence-based interventions in routine practice, while providing information on both the process and results of PA interventions – could provide the answer to questions related to feasible and effective translational research on PA promotion. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ypmed.2014.09.012. Conflict of interest statement The authors declare that they are no conflicts of interests.

Acknowledgments The project has received funding from a Network for Prevention and Health Promotion in Primary Care (redIAPP, RD12/0005) grant, research project grants (PI13/00573, PI12/02635, PI12/01914 and PS09/01461) from the Instituto de Salud Carlos III (Institute of Health Carlos III) of the Ministry of Economy and Competitiveness (Spain), co-financed with European Union ERDF funds, and Health Department of the Basque Government (EXP: 2009111072 and 2007111009). The authors would like to thank Edurne Zabaleta for the support and guidance provided in the present work and Elaine Lilly, Ph.D., for review of the English manuscript. References Campbell, F., Blank, L., Messina, J., et al., 2012. Physical Activity: Brief Advice for Adults in Primary Care. NICE Centre for Public Health Excellence.

Conn, V.S., Hafdahl, A.R., Mehr, D.R., 2011. Interventions to increase physical activity among healthy adults: meta-analysis of outcomes. Am. J. Public Health 101, 751–758. Eden, K.B., Orleans, C.T., Mulrow, C.D., Pender, N.J., Teutsch, S.M., 2002. Does counseling by clinicians improve physical activity? A summary of the evidence for the U.S. Preventive Services Task Force. Ann. Intern. Med. 137 (3), 208–215. Estabrooks, P.A., Glasgow, R.E., 2006. Translating effective clinic-based physical activity interventions into practice. Am. J. Prev. Med. 31 (4 Suppl), S45–S56. Estabrooks, P.A., Glasgow, R.E., Dzewaltowski, D.A., 2003. Physical activity promotion through primary care. JAMA 289, 2913–2916. Foster, C., Hillsdon, M., Thorogood, M., 2005. Interventions for promoting physical activity. Cochrane Database Syst. Rev. 1, CD003180. Grandes, G., Sánchez, A., Torcal, J., et al., 2008. Targeting physical activity promotion in general practice: characteristics of inactive patients and willingness to change. BMC Public Health 8, 172. Hallal, P.C., Andersen, L.B., Bull, F.C., et al., 2012. Global physical activity levels: surveillance progress, pitfalls, and prospects. Lancet 380, 247–257. Haskell, W.L., Lee, I.M., Pate, R.R., et al., 2007. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med. Sci. Sports Exerc. 39, 1423–1434. Hillsdon, M., 2013. Brief health professional-provided interventions may lead to small improvements in physical activity. Evid. Based Med. 18 (1), e1. Hillsdon, M., Foster, C., Thorogood, M., 2005. Interventions for promoting physical activity. Cochrane Database Syst. Rev. 1, CD003180. Hudon, C., Fortin, M., Soubhi, H., 2008. Single risk factor interventions to promote physical activity among patients with chronic diseases: systematic review. Can. Fam. Physician 54, 1130–1137. Lin, J.S., O'Connor, E., Whitlock, E.P., Beil, T.L., 2010. Behavioral counseling to promote physical activity and a healthful diet to prevent cardiovascular disease in adults: a systematic review for the U.S. Preventive Services Task Force. Ann. Intern. Med. 153, 736–750. Moyer, V.A., U.S. Preventive Services Task Force, 2012. Screening for and management of obesity in adults: U.S. Preventive Services Task Force recommendation statement. Ann. Intern. Med. 157, 373–378. Muller-Riemenschneider, F., Reinhold, T., Nocon, M., Willich, S.N., 2008. Long-term effectiveness of interventions promoting physical activity: a systematic review. Prev. Med. 47, 354–368. National Institute for Health and Care Excellence, 2013. Physical Activity: Brief Advice in Primary Care. NICE Public Health Guidance, London, p. 44 (guidance.nice.org.uk/ ph44). Orrow, G., Kinmonth, A.L., Sanderson, S., Sutton, S., 2012. Effectiveness of physical activity promotion based in primary care: systematic review and meta-analysis of randomised controlled trials. BMJ 344, e1389. Oxman, A.D., Guyatt, G.H., 1991. Validation of an index of the quality of review articles. J. Clin. Epidemiol. 44, 1271–1278. Royal Australian College of General Practice, 2012. Guidelines for Preventive Activities in General Practice, 8th ed. Royal Australian College of General Practitioners, East Melbourne. Scottish Intercollegiate Guidelines Network (SIGN), 2001. Management of Diabetes: A National Clinical Guideline. SIGN Publication Number 55. SIGN, Edinburgh. Smith, B., 2002. Do Primary Care Interventions to Promote PA Work? A Systematic Review of the Literature. The National Institute of Clinical Studies, Melbourne, Australia. Sørensen, J.B., Skovgaard, T., Puggaard, L., 2006. Exercise on prescription in general practice: a systematic review. Scand. J. Prim. Health Care 24 (2), 69–74. Starfield, B., 1998. Primary Care: Balancing Health Needs, Services and Technology, 2nd ed. Oxford University Press, New York. Stead, L.F., Buitrago, D., Preciado, N., Sanchez, G., Hartmann-Boyce, J., Lancaster, T., 2013. Physician advice for smoking cessation. Cochrane Database Syst. Rev. 5, CD000165. Tucker, J.M., Welk, G.J., Beyler, N.K., 2011. Physical activity in U.S.: adults compliance with the Physical Activity Guidelines for Americans. Am. J. Prev. Med. 40, 454–461. Tulloch, H., Fortier, M., Hogg, W., 2006. Physical activity counseling in primary care: who has and who should be counseling? Patient Educ. Couns. 64 (1–3), 6–20. van Doorslaer, E., Masseria, C., Koolman, X., OECD Health Equity Research Group, 2006. Inequalities in access to medical care by income in developed countries. CMAJ 174, 177–183. van Sluijs, E.M., van Poppel, M.N., van Mechelen, W., 2004. Stage-based lifestyle interventions in primary care: are they effective? Am. J. Prev. Med. 26, 330–343.

Please cite this article as: Sanchez, A., et al., Effectiveness of physical activity promotion interventions in primary care: A review of reviews, Prev. Med. (2014), http://dx.doi.org/10.1016/j.ypmed.2014.09.012