Effectiveness of health literacy- and patient activation-targeted interventions on chronic disease self-management outcomes in outpatient settings: a systematic review

Keywords: patient activation, health literacy, chronic diseases, self-management, outpatient settings, health related-outcomes, patient related outcomes, behavioural outcomes.

The prevalence of chronic disease has been increasing exponentially worldwide over the past decades due to the aging population and the availability of better diagnostic tools (Hajat and Stein 2018). Globally, the number of people aged >60 years increased from 382 million in 1980 to 962 million in 2017, and this is expected to double in 2050 (United Nations 2017). In 2019, non-communicable chronic diseases such as ischaemic heart disease, stroke, lung cancer, depression and diabetes were the top conditions contributing to the highest percentage of morbidity and mortality worldwide (Lujic et al. 2017; Hajat and Stein 2018; Niknami et al. 2018; Jeganathan and Hosseinzadeh 2020). This has resulted in increasing healthcare costs due to health care utilisation, low productivity and negative effects on individuals’ quality of life (Edwards and Hosseinzadeh 2018; Hajat and Stein 2018; Dahal and Hosseinzadeh 2019; Almutairi et al. 2020; Ansari et al. 2021). Globally, the economic burden of chronic diseases over the next two decades is estimated to cost approximately US$47 trillion (Bloom et al. 2011).

Self-management has been adopted as a key strategy in the management of chronic diseases (Yang et al. 2017). The available evidence suggests that improving disease knowledge and adopting healthy behaviours, such as increasing physical activity levels, medication adherence and smoking cessation, have been associated with better outcomes, including improving patients’ quality of life and reducing health care-associated expenditure (Dadich and Hosseinzadeh 2013; Yang et al. 2017; Ho et al. 2018; Ansari et al. 2020; Hosseinzadeh et al. 2020).

Patient activation (PA) and health literacy (HL) are associated with improved self-management outcomes (Smith et al. 2013; Hibbard 2017). PA refers to having the knowledge, skills and confidence to manage a health condition. Highly activated patients tend to have an active role in their disease management through shared decision making, goal setting and involvement in health behaviours change (Hibbard and Greene 2013; Greene et al. 2015).

HL has been defined at three levels: functional, communicative and critical HL (Nutbeam 2008). Functional HL includes basic literacy skills in reading and writing, communicative HL includes the cognitive skills that enable a patient to access and navigate the healthcare system and ensure effective communication with healthcare professionals and critical HL includes advanced social and cognitive skills that enable the patient to appraise and integrate health information to have control over their health (Nutbeam 2008). Patients with low HL are more likely to have poorer outcomes due to limited understanding of disease management (Kale et al. 2015). Low HL also leads to limited access to available supportive services and poor communication with healthcare providers (Omachi et al. 2013).

The literature suggests that targeting both PA and HL in self-management interventions could result in more promising health outcomes and behaviour change than using PA or HL alone (Smith et al. 2013; Yadav et al. 2019). However, there are inconsistent findings about the effectiveness of such interventions on chronic disease self-management outcomes. The aim of this systematic review is to fill this gap and summarise the available evidence from randomized control trials (RCTs) that evaluated the effectiveness of PA- and HL-driven self-management interventions to improve health-related outcomes among patients with chronic diseases in outpatient settings.

Six well known databases (MEDLINE, Web of Science, Scopus, Science Citation Index, EMBASE and Academic Search Complete) were searched. The search aimed to identify RCTs that assessed both PA- and HL-driven chronic diseases self-management interventions published since January 2004 until June 2021. The literature was reviewed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol (Moher et al. 2009; Fig. 1; Box 1). Search terms included chronic disease, health literacy/health education, patient activation/engagement, self-management/self-care and primary care/outpatient clinic/community setting (Fig. 1). This review was not registered.

Fig. 1.  Flow chart for the study search and selection criteria.

A logic model was developed from reviewing the included articles (Fig. 2).

Fig. 2.  Logic model of HL and PA targeted chronic disease self-management interventions.

RCTs were only eligible for inclusion if they had appropriate measures of PA and HL, used a PA- and HL-driven self-management intervention, were peer reviewed and were published in English (Table 1). The trials were reviewed independently by all three authors. Any disagreements were resolved by discussion. Data were extracted and tabulated by one author (MS) and double-checked by the other two authors (HH, SD).

Table 1.  Inclusion and exclusion criteria

The Cochrane Back Review Group (CBRG) assessment criteria were used to assess the quality of the trials reviewed. Only trials scoring ≥6 CBRG points were included in the present study. The included interventions were reviewed using the PRISMA protocol (Moher et al. 2009). Narrative synthesis was used to analyse the literature due to heterogeneity of the chronic diseases, the interventions and the outcomes measured.

The literature search identified 271 potential studies; of these 123 (duplicates) and 99 (not an RCT or not measuring PA and HL) were excluded. The titles and abstracts of the remaining 49 studies were reviewed, and only 21 studies met the selection criteria. The full text of these 21 studies was further assessed, with 10 studies finally included in the review (Table 2).

Table 2.  Characteristics of the studies eligible for inclusion in this review
AADES, American Association of Diabetes Education Standards; AUDIT-C, Alcohol Consumption Questions of the Alcohol Use Disorders Identification Test; BP, blood pressure; BS, blood sugar; CAT, COPD Assessment Test; CHWs, community health workers; COPD-Q, Chronic Obstructive Pulmonary Disease knowledge Questionnaire; CVD, cardiovascular disease; DSMP, diabetes self-management program; ePHR, electronic personal health record; EQ-5D, EuroQol-5 Dimension; EQ-5D-3L, EuroQol-5 Dimension –3 levels; FCCHL, Functional Communicative Critical Health Literacy; FFKA, Freiburg Questionnaire for Physical Activity; HADS, Hospital Anxiety and Depression Scale; HF, heart failure; MARS-D, Medication Adherence Report Scale; JNC7, Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure; MET, metabolic equivalent; MI, motivational interviewing; PAM-13, 13-item patient activation measure; Pam AM300, personal activity monitor; PHQ-9, Patient Health Questionnaire-9; PRAISE, Pulmonary Rehabilitation Adapted Index of Self-Efficacy; QoL, quality of life; RHFSCBS, Revised Heart Failure Self-Care Behaviour Scale; SCHFI, Self-Care of Heart Failure Index; SDM, shared decision-making; SDSCA, Summary of Diabetes Self-care Activities; SeMas, Self-Management Screening; SF-12, Short Form 12 Health Survey; SM, self-management; SM-HL, social media health literacy; SOC, Stages of Change Across 10 Health Risk Behaviors for Older Adults; T2D, Type 2 diabetes; TEL-HL, telephone-based health literacy; VAS, visual analogue scale

The ratio of males to females ranged from 24% to 91%, with a higher proportion of females in most studies. The age of the participants ranged between 30 and 84 years, and the sample size ranged from 60 and 4283 patients (Table 2).

The studies included scored ≥6 points on the CBRG assessment criteria (Furlan et al. 2009), which indicates a moderate or high quality of the study and a low risk of bias. Most studies were single blinded.

Components of the interventions are described in Table 2. The reviewed studies targeted a wide range of chronic diseases, including hypertension (Ryvicker et al. 2013), diabetes (Lorig et al. 2009; Kim and Utz 2019), heart failure (Young et al. 2016), chronic obstructive pulmonary disease (COPD; Eikelenboom et al. 2016; Chang and Dai 2019) and HIV (Carroll et al. 2019). HL was assessed using outcome measures such as functional, communicative and critical outcomes, social support and disease knowledge. PA was assessed using the Patient Activation Measure (PAM; Hibbard et al. 2005). Several strategies and techniques have been used to tackle patients’ low PA and HL levels. The strategies and techniques applied included motivational interviewing (Ryvicker et al. 2013; Dwinger et al. 2020), collaborative goal setting (Lorig et al. 2009; Chang and Dai 2019; Westland et al. 2020), building patients’ skills to encourage shared decision making (Lorig et al. 2009; Carroll et al. 2019) and self-management support, either through the use of social media and/or telephone (Young et al. 2016; Kim and Utz 2019; Dwinger et al. 2020) or through communication with a community health worker (Kangovi et al. 2018).

Most of the interventions were delivered by nurses (Ryvicker et al. 2013; Eikelenboom et al. 2016; Young et al. 2016; Chang and Dai 2019; Kim and Utz 2019; Dwinger et al. 2020; Westland et al. 2020), but some were delivered by peer coaches (Lorig et al. 2009; Carroll et al. 2019), one was delivered by collaboration between peer coaches and facilitated nurses (Carroll et al. 2019) and one was delivered by a community health worker (Kangovi et al. 2018). The duration of the interventions ranged from 3 months to 4 years.

The impact of the interventions was reviewed using four outcome domains: patient-related outcomes such as HL and PA and self-efficacy; behavioural outcomes such as physical activity; and health-related outcomes, including quality of life, mental health and health care use outcomes. Table 2 describes the measured outcomes and Table 3 summarises the findings of the studies.

Table 3.  Summary of outcomes results
BMI, body mass index; BNP, B-type natriuretic peptide; BP, blood pressure; CHW, community health worker; CI, confidence interval; COPD, chronic obstructive pulmonary disease; DSMP, diabetes self-management program; EQ-5D, EuroQol-5 Dimension; EQ-5D-3L, EuroQol-5 Dimension –3 levels; EQ-VAS, EuroQol Visual Analogue Scale; PATCH, Patient AcTivated Care at Home; QoL, quality of life; SM, self-management; SM-HL, social media health literacy; TEL-HL, telephone-based health literacy

Functional, communicative and critical HL outcomes were used across the studies. The impact of HL on patients’ understanding of their disease was assessed by two studies (Young et al. 2016; Chang and Dai 2019), but only one of them reported a significant improvements in patients’ disease knowledge (Chang and Dai 2019). Similarly, the effect of interactive HL on physician communication was assessed in two studies (Lorig et al. 2009; Carroll et al. 2019), but only one reported significant improvements in physician communication (Lorig et al. 2009). Critical HL was assessed by one study that found significant improvement among patients in the intervention group (Kangovi et al. 2018).

PA levels in all 10 trials were assessed using the PAM (Hibbard et al. 2005), with six studies reporting improvements in PA levels (Lorig et al. 2009; Young et al. 2016; Carroll et al. 2019; Chang and Dai 2019; Kim and Utz 2019; Dwinger et al. 2020). Five studies assessed self-efficacy (Lorig et al. 2009; Young et al. 2016; Carroll et al. 2019; Chang and Dai 2019; Westland et al. 2020). The changes in self-efficacy were consistent with changes in PA levels in four studies (Lorig et al. 2009; Young et al. 2016; Chang and Dai 2019; Westland et al. 2020); three of these studies showed improvement in both PA and self-efficacy (Lorig et al. 2009; Young et al. 2016; Chang and Dai 2019), whereas Westland et al. (2020) failed to find improvements in either measure.

Five studies assessed patients’ physical activity levels (Lorig et al. 2009; Eikelenboom et al. 2016; Young et al. 2016; Dwinger et al. 2020; Westland et al. 2020); of these, three reported improvements in self-reported physical activity levels (Lorig et al. 2009; Young et al. 2016; Dwinger et al. 2020). PA was found to be the major mediator of improvement in self-management behaviours, including physical activity levels, especially among patients with low levels of disease knowledge (Young et al. 2016). Lorig et al. (2009) reported significant improvements in PA leading to slight improvements in aerobic exercise among patients in the intervention group, and Westland et al. (2020) found that patients with low perceived social support reported improvements in their physical activity levels.

Three studies assessed mental health outcomes, such as depression and/or anxiety (Lorig et al. 2009; Kangovi et al. 2018; Dwinger et al. 2020). Of these, Lorig et al. (2009) found significant improvements in symptoms of depression and Kangovi et al. (2018) also reported improvements in patients’ mental health.

Five studies assessed quality of life parameters (Kangovi et al. 2018; Carroll et al. 2019; Chang and Dai 2019; Dwinger et al. 2020; Westland et al. 2020). None of the studies found improvements in quality of life after the interventions.

Four studies assessed health care utilisation (Lorig et al. 2009; Young et al. 2016; Kangovi et al. 2018; Chang and Dai 2019), with none finding improvements after the interventions. Unexpectedly, Young et al. (2016) reported a significant increase in hospitalisations among patients in the intervention group at the 6-months follow-up; however, this was reduced to insignificant levels at the 12-month follow-up compared with the control group.

Three studies reported significant positive associations between PA and HL levels (Ryvicker et al. 2013; Eikelenboom et al. 2016; Kim and Utz 2019), whereby Kim and Utz (2019) found that patients with low HL at baseline were more likely to have higher PA levels at the 9-week follow-up. In addition, five studies evaluated the association between PA levels and self-management outcomes (Ryvicker et al. 2013; Young et al. 2016; Carroll et al. 2019; Kim and Utz 2019; Westland et al. 2020). Patients with low PA levels showed the most benefit from the intervention compared with those who had high PA levels at baseline (Ryvicker et al. 2013; Carroll et al. 2019; Westland et al. 2020).

This review aimed to summarise the available literature on the effectiveness of PA and HL on chronic disease self-management in outpatient settings.

The studies reviewed varied in many aspects, such as type of chronic disease, intervention components, delivery mode, follow-up duration, sample size and outcome variables. Some studies focused on social support to motivate patients and tackle low HL and PA levels (Dwinger et al. 2020), whereas others focused on other techniques, such as motivational interviewing (Ryvicker et al. 2013; Young et al. 2016; Kangovi et al. 2018; Kim and Utz 2019; Dwinger et al. 2020) and building patient skills, such as problem solving and asking questions to improve patient–provider communications (Lorig et al. 2009; Carroll et al. 2019). Many studies focused on using instructions appropriate for patients with low HL levels. Due to these wide variations, the results will be interpreted with some caution.

The reviewed studies evaluated interactions between PA and HL (Ryvicker et al. 2013; Eikelenboom et al. 2016; Carroll et al. 2019; Kim and Utz 2019; Westland et al. 2020) and their effects on chronic disease self-management outcomes (Lorig et al. 2009; Young et al. 2016; Kangovi et al. 2018). Most of the studies included showed improvements in patients’ HL outcomes, including HL levels (Carroll et al. 2019; Dwinger et al. 2020), disease knowledge (Chang and Dai 2019), communication with healthcare providers (Lorig et al. 2009) and social support (Kangovi et al. 2018; Westland et al. 2020). Some studies also reported improvements in PA levels (Lorig et al. 2009; Young et al. 2016; Carroll et al. 2019; Chang and Dai 2019; Kim and Utz 2019; Dwinger et al. 2020). Improvements in HL and PA resulted in better physical activity levels (Young et al. 2016; Dwinger et al. 2020) and mental health status (Lorig et al. 2009; Kangovi et al. 2018). However, these benefits were not translated into improvements in quality of life (Kangovi et al. 2018; Carroll et al. 2019; Chang and Dai 2019; Dwinger et al. 2020; Westland et al. 2020) and healthcare utilisation (Lorig et al. 2009; Young et al. 2016; Kangovi et al. 2018; Chang and Dai 2019). This is in line with the findings from previous systematic reviews, which concluded that improving PA and HL results in changes in patients’ behaviours (Taggart et al. 2012; Lundell et al. 2015).

Nevertheless, the evaluation of the interaction of PA and HL with self-management outcomes found a positive association between self-management outcomes and PA and HL levels (Ryvicker et al. 2013; Eikelenboom et al. 2016; Carroll et al. 2019; Kim and Utz 2019; Westland et al. 2020), in that patients with low PA and HL levels benefited most from the interventions. This finding is consistent with the currently emerging evidence that suggests that PA and HL are essential factors to ensure the success of self-management and should be combined to further achieve improvements in long-term chronic disease self-management outcomes (Smith et al. 2013; Yadav et al. 2019, 2020).

Our review further revealed that the interventions were not associated with improved quality of life and hospital admissions. This could be explained, in part, by the inability of some interventions to successfully change patients’ activation levels across the range of baseline PA levels. Thus, this highlights the importance of providing intervention tailored to patients’ PA levels (Adams 2010). This concurs with our previous systematic review among COPD patients, which suggested that changing self-efficacy is central to bringing about positive changes in end-point outcomes (Hosseinzadeh and Shnaigat 2019). This further highlights the need for long-term interventions to support patients and achieve consistent and long-term benefits (Hosseinzadeh and Shnaigat 2019).

To conclude, this review suggests that both HL and PA have a major role in improving chronic diseases self-management intervention outcomes. Future studies should address patients’ PA and HL levels to achieve long-term outcomes.

This review has some limitations, including the inclusion of different chronic diseases, which could have a direct effect on the measured outcomes. In addition, conceptualisation of HL was different in each RCT, and this is an issue for comparability. Some studies focused on patients, whereas others were focused on the practitioner and social levels. Many studies were based in US, which may limit the generalisability of the results to other countries. In addition, the populations of some studies were not representative of the wider population, because some studies had related to minority only and others had high minority representations. Furthermore, there was variability in the follow-up duration between studies.

HL- and PA-targeted chronic diseases self-management has yielded moderate improvements in HL, PA and self-efficacy levels. There were some improvements in physical activity and mental health outcomes, and, interestingly, patients with low PA levels gained the most benefits from the interventions. Future studies should focus on using PA- and HL-tailored interventions to further improve outcomes for patients with chronic disease. Furthermore, there is a need to consider a comprehensive HL tool measuring functional, communicative and critical HL to obtain conclusive outcomes.

Data sharing is not applicable because no new data were generated or analysed during this study.

The authors declare that they have no conflicts of interest.

This research did not receive any specific funding.