An Integrative Review of Nurse-Delivered Physical Activity Interventions in Primary Care

Purpose:

Promotion of physical activity (PA)has been a public health priority for decades (U.S. Department of Health and Human Services, 2010). Participation in regular PA reduces the risks of several chronic diseases, cancers, and all-cause mortality (Physical Activity Guidelines Committee [PAGC], 2008). Despite the substantial health benefits of being physically active, most Americans do not meet guidelines (CDC, 2013).

Eighty-percent of U.S. adults see a healthcare provider at least once a year (Blackwell, 2014) with 56% of all office visits in the primary care setting, making this a valuable location for health promotion (Blackwell, 2014).  Previous reviews of primary care PA interventions have shown promising results (Garrett, 2011; Orrow, 2012). However,  non-adherence to disease prevention recommendations, including increased PA, has been found to be up to 80% (Christensen, 2004). Nurses are uniquely positioned to help motivate patients to make health behavior changes such as through PA counseling.  The specific role of nurses in PA promotion in primary care is understudied.  The purpose of this integrative review is to describe nurse-delivered physical activity interventions conducted in primary care settings and determine intervention attributes which led to positive changes in PA.

 Methods:

Design

An integrative review was conducted to synthesize the peer-reviewed literature to describe the roles of nurses in PA interventions conducted in primary care settings. Whittemore and Knafl’s (2005) suggested methodology for integrative reviews was followed. 

Search Methods

A systematic search of existing peer-reviewed physical activity intervention studies conducted by nurses in the primary care setting was conducted through the following online databases: CINAHL, PubMed, PsycINFO, SportDiscus, Cochrnae, and Sigma Theta Tau Research Repository. Key words used in the search included physical activity or exercise or walking and primary health care.

Results:

Search Outcome

The initial computerized search yielded 1,736 citations. After removing duplicates and screening titles and abstracts, 392 potentially relevant studies were identified for evaluation.  After full-text readings, 375 were excluded. Four additional studies were identified through ancestry searches on previously published review articles and all potential primary studies. The final sample included 19 articles.

Sample Characteristics

Seven of the 19 studies were conducted in the United States. The 19 studies represented a total of 7,350 participants with a range of ages from 18-91years (mean ages 28-72 years). Sample sizes ranged from 20-1173. Gender was reported in 16 studies with a range of % male from 0 to 100%, with an average of 40%. Six studies reported racial ethnicity with a range of 8-30% non-Caucasian or non-white European.

Methodological Attributes

Fifteen of the 19 studies included control groups of subjects who received usual primary care.  Half (n=10) of the interventions were guided by health behavior theoretical frameworks. Aspects of the transtheoretical model, specifically stages of change, were used in seven of the studies reviewed. Other theories used to guide the interventions were used in seven of the studies. These included social cognitive theory, mainly focusing on self-efficacy (Dubbert, Morey, Kirchner, Meydrech, & Grothe, 2008; Nanette Mutrie et al., 2012; Naylor, Simmonds, Riddoch, Velleman, & Turton, 1999), theory of planned behavior (Little et al., 2004), PRECEDE-PROCEED Model (Kinnunen et al., 2007), and the adapted physical activity model (Purath, Keller, McPherson, & Ainsworth, 2013).

Fourteen studies used self-report PA measures collected by questionnaires (n=11) or logbooks (n=3), while only six studies used objective data collected by pedometer or accelerometer (McKay et al., 2009; Nanette Mutrie et al., 2012; Piette et al., 2011; Bonnie J. Sherman, Gina Gilliland, Jeanne L. Speckman, & Karen M. Freund, 2007; Verwey et al., 2014).

The timing of follow-up measurement of outcomes varied across the studies, with the duration of follow-up in 16 studies ranging from 3 to 12 months. One study had follow-up data collection at 1 month (Little et al., 2004); two other studies had follow-ups extending to 2 years (Lawton et al., 2009) and 4 years (Murchie et al., 2003).

Intervention Attributes

Intervention delivery modes included traditional face-to-face, telephone, mail and web-based deliveries. Eleven studies used single delivery mode: face-to-face delivery (n=9), telephone (n=1), or web-based delivery (n=1). The remaining eight studies used a combination of delivery modes including face-to-face and telephone deliveries (n=6), or face-to-face, telephone and mail deliveries (n=2). 

The intervention dose in three of the studies consisted of a one-time counselling or education session for the intervention group (Elley, Kerse, Arroll, & Robinson, 2003; Naylor et al., 1999). In eight other studies, an initial counselling or education session was supported with 1 to 4 more sessions over 3 to12 months (Dubbert et al., 2008; Kinnunen et al., 2007; Leonhardt et al., 2008; Murchie et al., 2003; B. J. Sherman, G. Gilliland, J. L. Speckman, & K. M. Freund, 2007; Steptoe et al., 1999; van Sluijs et al., 2005; Verwey et al., 2014). One study provided 21 sessions of cognitive behavior therapy with subjects over a 12 month period (Piette et al., 2011). Supportive or motivational follow-up contacts were integrated with counselling sessions in four studies (Elley et al., 2003; Murchie et al., 2003; Steptoe et al., 1999; van Sluijs et al., 2005), while in another study, motivational contacts were delivered alone (Rhudy, Dubbert, Kirchner, & Williams, 2007). Three studies conducted pedometer-based walking programs over a 3 or 6 month period (McKay et al., 2009; Nanette Mutrie et al., 2012; Bonnie J. Sherman et al., 2007). The remaining three studies provided exercise prescriptions with or without follow-up contacts over 3 to 9 months (L. Josyula & R. Lyle, 2013; Lawton et al., 2009; Purath et al., 2013). 

Role of Nurse

The majority of studies (n=18) involved only nurse or nurse practitioner in intervention delivery, except for one study which involved both nurse and exercise specialists (Elley et al., 2003). Nurse practitioners provided exercise prescriptions and/or counselling in seven studies. The role of nurses varied in studies and included leading exercise training or monitoring programs, PA counselling, and providing motivational telephone calls. In device-based exercise programs, nurses distributed activity monitors and provided supportive contacts as well.

Efficacy of Interventions for Increasing Physical Activity

Fifteen of the 19 studies reported greater PA in intervention subjects than in control subjects (del Rey-Moya et al., 2013; Dubbert et al., 2008; Elley et al., 2003; L. K. Josyula & R. M. Lyle, 2013; Lawton et al., 2009; Little et al., 2004; McKay et al., 2009; Murchie et al., 2003; N. Mutrie et al., 2012; Piette et al., 2011; Purath et al., 2013; Rhudy et al., 2007; B. J. Sherman et al., 2007; Steptoe et al., 1999; Verwey et al., 2014). While most studies demonstrated efficacy, PA measures were usually self-reported, collected by questionnaires or logbooks (n=9), except for six studies collecting data by PA monitors. Findings in four of the 19 studies did not demonstrate significant difference between intervention and control groups in levels of walking or overall PA.

 Conclusion:

The most common nurse-delivered intervention was PA counselling with supportive or motivational contacts; others were exercise training or monitoring, device-based exercise program and exercise prescriptions. As previously stated, four of the studies did not yield significant intervention effects on PA outcomes. These outcomes may have been attributed to sufficient levels of PA at baseline (Leonhardt et al., 2008), underlying health conditions (Kinnunen et al., 2007; Leonhardt et al., 2008), or low dose of intervention (Leonhardt et al., 2008; Naylor et al., 1999).

Several of the studies included in this review have important limitations which should be addressed in future research. Of the 19 studies reviewed, only half of the interventions were designed around a health behavior theory.  Furthermore, of those interventions which were guided by theory, most only incorporated one or two aspects of the theory instead of the theory as a whole.  Future interventions studies should explicitly state the theoretical foundation and utilize the theory as a whole. This would strengthen the contribution of individual studies to conceptual understandings of health behavior change.

Furthermore, almost 74% of included studies relied upon self-reported PA to measure outcomes.  Self-report measures are prone to recall and social desirability bias. Objective assessments of PA with the use of pedometers or accelerometers should be incorporated in future interventional research.  In addition, future research should follow participants one year and beyond to determine effects on maintenance of PA behavior change.

No studies examined the cost-effectiveness of the interventions provided.  This is an important limitation of current studies as a commonly reported barrier to PA promotion among medical professionals is the lack of reimbursement for PA counseling (McPhail & Schippers, 2012). Care provided by nurses and nurse practitioners has been proven to be more cost effective than physicians (Chenoweth, Martin, Pankowski, & Raymond, 2008; Coddington & Sands, 2008). Therefore, utilizing nurses and nurse practitioners to deliver PA interventions may be a cost-effective way to increase population-levels of PA.

Findings from this integrative review indicate that nurse-delivered PA interventions in primary care show overall effectiveness in increasing PA level in general population. Future rigorously designed studies are warranted to guide PA promotion in primary care.