Background

The World Health Organization (WHO) has targeted a reduction of 30% in the number of premature deaths due to chronic disease by 2030 [1] and highlighted the importance of patient empowerment in prevention and health promotion efforts. The main idea is to redistribute power from the health professionals to the patients who handle the challenges of chronic disease on a daily basis [2].

Within healthcare, the empowerment approach is a patient-centric, collaborative approach that starts with the principle of individuals’ inherent capacity to be in control of one’s own life. It has been described as a complex experience of personal change, facilitated by healthcare providers [3]. A major concern of people diagnosed with chronic disease is the multitudes of uncertainties they face, and the restrictions of their life space they encounter due to their health condition [4, 5]. This uncertainty may be accompanied by the experience of lack of control and the feeling of powerlessness that is connected to the disruption of the physical, psychological and social aspects of the patients’ lives [6,7,8]. Patients frequently undergo long periods of multimodal treatment and are challenged to change their lifestyle in order to prevent their chronic condition from worsening. However, patients possess internal and external strengths and self-management strategies to fight against the threat of deterioration while living with chronic disease [9]. These capacities function to empower patients to regain control through a process of health-related change [5, 10]. According to Castro, Regenmortel, et al. [11] “patients have come to be seen as experts of their own bodies, symptoms and situations, and patients’ experiential knowledge is now considered to be complementary to professional knowledge and important for the success of their treatment, self-care management, and for improving the quality of care” (p.1924). The challenge is to effectively utilize patient empowerment resources [6] and capabilities to promote health and wellness within chronic disease. Empowerment is an essential concept in this regard. A descriptive review [12] of 49 empirical studies showed that 35 different definitions were used to define empowerment and that the use of the concept has been inconsistent. Furthermore 38 different instruments were applied to measure empowerment. The lack of clear conceptualization of empowerment as well as the interchangeably use of empowerment with other related concepts, suggests the need to distinguish between empowerment as a psychosocial and health-related process, and self-management, self-efficacy, patient activation, health literacy, behavioral change and quality of life as indicators or outcomes of empowerment [12,13,14]. In this review, we have chosen to rely on WHO’s definition of the empowerment process as this implies a broad understanding of empowerment that allows for the inclusion of different definitions and interventions covering a variety of chronic disease diagnostic categories. WHO has defined four fundamental components of the empowerment process: patient participation, patient knowledge, patient skills, and the creation of a facilitating environment and have integrated empowerment in their guidelines for reaching sustainable goals [15,16,17]. A facilitating environment implies being listened to with regard to one’s concern, being engaged in shared decision-making with health care professionals and having access to high quality organized care, i.e. interventions for knowledge improvement and psychological support [18]. The WHO has published a handbook in 2021 [17] that focuses on empowerment and reaffirms the importance of social participation as fundamental for empowerment.

A wide variety of interventions that facilitate partnerships between patients and healthcare professionals (HCP) using a wide-range of approaches including patient education, shared decision-making, goal-setting, on self-evaluation, and motivational interviewing have been developed [19,20,21]. Cardoso Barbosa, de Queiroz Oliveira, et al. [22] integrative review showed that empowerment interventions have the potential to strengthen patient autonomy and the trust of individuals in shared decision-making, helping them to develop coping and communication skills, and implementing behavior changes related to their disease, underscoring the importance of focus on empowerment in interventions for living a good life with a chronic condition.

The focus of our review is to study empowerment interventions for patients with chronic disease across diagnostic categories. In a prior review, Chen and I-Chuan [23] demonstrated that empowerment-focused interventions improved the health status, psychological condition and quality of life (QOL) of chronically ill patients. Werbrouk, Swinnen, et al. [3]’s systematic review covering 2007–2017 included 32 randomized-controlled trials (RCT) of which 23 were included in a meta-analysis that estimated an overall interventional effect but with high heterogeneity. Samoocha, Bruinvels, et al. [24] studied the effect of web-based interventions for patient empowerment. The included Web-based interventions had a positive effect on empowerment in diabetes patients, on self-efficacy, however when compared to face-to-face delivery interventions, no significant effects were found for mastery. This 2010 review has not been updated, however, more recent systematic reviews involving patients with specific chronic conditions shows similar results. For example, a systematic review on web-based interventions targeting cardiovascular risk factors in older adults [25], found a potential to improve the cardiovascular risk profile, however, the effects were modest and declined with time. Another review assessing the effectiveness of internet empowerment-based self-management interventions within metabolic diseases [26] showed more positive results. These interventions significantly improved the health status of adults with metabolic diseases, in particular their exercise habits, HbA1c levels, body weight, empowerment and quality of life. The majority of studies were limited to patients with diabetes and had methodological issues with increasing risk of bias. When searching the Cochrane review database for empowerment interventions, we found that the majority of former reviews were disease-specific [27, 28]. Two reviews of interventions in chronic illness in general were found, however, these reviews did not mention empowerment. De Jongh, Gurol‐Urganci, et al. [29] reviewed interventions using mobile phone messages to facilitate self-management and found no statistical difference in health outcomes from text messages compared with usual care, however, moderate improvement in self-management capacity was found in diabetes patients. Smith, Wallace, et al. [30] reviewed the effectiveness of interventions designed to improve clinical and mental health outcomes and patient reported outcomes in people with multimorbidity in primary care and community settings and found no clear positive improvements, however, modest improvement was found in health outcomes among patients with depression.

The results from former reviews show mixed results while the efficacy of empowerment interventions for patients with different kinds of diagnosis, remains unclear. As former research has identified a wide variety of conceptual understandings of empowerment, there is a need to clarify whether interventions capture essential components of empowerment as described by WHO [15]. In response, our objective was to conduct a systematic review of empowerment interventions that covered a broad range of chronic diseases in order to assess if fundamental components of empowerment were included and to study intervention efficacy across diagnoses.

Method

This systematic review and meta-analysis aimed to describe the characteristics of empowerment interventions covering a broad range of chronic diseases by a) summarizing intervention setting, structure, modality, content, and clarify consistency with WHO’s empowerment definition, b) to summarize outcome measures and estimate the effect in group and individual formats.

The study was designed as a systematic literature review and meta-analysis covering January 1st 2016-March 25th 2020 and reported according to PRISMA guidelines [31]. The protocol was published in PROSPERO (CRD42020178286).

Search methods

The research question was structured according to PICOs (Population, Intervention, Comparison, Outcome, and Study design) (Table 1). Eligibility criteria was: a) scientific publication of original research, b) RCT/quasi experiments and c) included patients with different chronic diseases.

Table 1 Inclusion and exclusion criteria in PICOS (Population, Intervention, Comparison, Outcome, Study Design) format
Full size table

A systematic search strategy was implemented in MEDLINE, Embase, PsycINFO, CINAHL, Web of Science, and COCHRANE Central Register of Controlled Trials (Supplementary file 1) and was reviewed using the Peer Review of Electronic Search Strategies (PRESS) checklist [32]. Controlled vocabulary (e.g., Medical Subject Headings in MEDLINE) and additional keywords were used to identify relevant search terms, and the RCT filter from Ovid Expert searches was applied from which time the prior systematic review by Chen and I-Chuan [23] had been completed. Reports written in English and Scandinavian languages were included.

Search outcomes

Search results were imported to an EndNote library, and duplicates were removed. Reports were randomly divided in two halves and imported into Rayyan software for review by two teams of reviewers (X & X, X & X). The abstracts and titles were independently screened by two reviewers in accordance with the eligibility criteria. The two reviewers then retrieved and screened the full texts of the relevant studies, then reviewer pairs evaluated the alternate set of reports from the first screening. Disagreements about inclusion were solved by discussing with the whole team. The second screening resulted in 83 studies with significant differences in scope regarding interventions, populations and outcomes. In order to be able to do an in-depth description of the interventions, a pragmatic decision was made to exclude studies published prior to 2016. This choice resulted in the exclusion of 44 publications, of which 57% described patients with diabetes. Reviews covering empowerment interventions in diabetes are previously published [33,34,35]. A PRISMA flow chart was used to document the number of excluded studies and the reasons for exclusion.

Quality appraisal

We applied the updated Cochrane Risk of Bias Assessment tool (ROB2). Reports were classified as having a low risk of bias, some concerns, or a high risk of bias [36]. All quality assessments were independently conducted in the review pairs and consensus was achieved through discussion among the whole team.

Data extraction

The following data was extracted: author, year of publication, setting, patient diagnosis, type of health care professional delivering the intervention, modality, design, outcome measures and results (Table 2). Extracted data were cross-checked and finalized by all team members.

Table 2 Summary of individual studies
Full size table

Synthesis

Meta-analyses were performed using RevMan V5.2 software. To summarize continuous data, the pooled mean difference (MD) and 95% confidence interval (CI) were calculated. Given the included interventions were delivered in different modes, formats, sessions and duration, random-effect models were used in the pooled analysis [81]. The I2 metric describes the percentage of total variation across studies due to heterogeneity. The Q-value was used to examine the degree of heterogeneity.

Results

This systematic review yielded 39 empowerment-focused intervention studies conducted among 8,011 participants. We retrieved 2,233 reports after the removal of duplicates and excluded 1,992 reports based on the title, abstract, and keywords, leaving 241 reports that were assessed for eligibility by reviewing the full-text (Fig. 1).

Fig. 1
figure 1

PRISMA flow chart. The figure details our search and selection process applied during the systematic review

Full size image

Characteristics of the studies included

The characteristics of the individual studies are summarized in Table 2. Of 39 included studies, 32 (85%) were RCTs [19, 21, 37, 38, 41, 43, 45, 46, 49, 52,53,54, 57,58,59,60,61,62,63,64, 66, 68,69,70,71, 74,75,76,77,78,79,80]. Seven studies used other methods including five quasi-randomized trials, [39, 56, 65, 72, 73] one pre-post design, [51] and one partially randomized patient preference trial [48]. All studies were published in English within the 5-year period spanning 2016–2020, inclusive.

A total of 15 different countries were represented in the review. Most studies were conducted in Iran (n = 13), followed by Brazil (n = 4), China (n = 4), Netherlands (n = 3), United States (n = 2) and Turkey (n = 2). One study was conducted in each of Denmark, Denmark/Turkey, Greece, Hong Kong, Malaysia, Portugal, Qatar, South Korea, Taiwan, and Thailand (see Table 2).

The studies included a variety of chronic diseases with most conducted among patients with a diagnosis of Type 2 diabetes mellitus (n = 22) followed by cardiovascular disease (n = 3), stroke (n = 3), chronic obstructive pulmonary disease (COPD) (n = 2), heart failure (n = 2) cancer (n = 3). One study included patients who had undergone a renal transplant, one was conducted among veterans with depression and diabetes, one study included patients suffering from chronic anxiety and depression. In addition, one study included patients with at least one officially diagnosed chronic disease. Mean age ranged between 46.9 (SD 5.5) [52] and 73.29 years (SD 8.6) [69].

The types of health care professionals (HCP) involved in intervention delivery varied to a large degree across studies: 22/39 studies included nurses, 8/39 studies used multidisciplinary teams, 3/39 studies used physicians, and one study each with dentists, research psychologists, and a research team. Four studies were unclear about the types of HCP who were involved; [46, 51, 64, 79].

Retention rate of the intervention groups ranged between 38 and 100%. Retention rate of the control groups ranged between 46 and 100%.

Intervention characteristics

Intervention settings, modalities and content varied across studies. Settings included outpatient or community clinics (n = 32), inpatient settings (n = 1) and mixed settings (n = 5) where patients began the intervention as inpatient and continued post-discharge in the outpatient setting. One study did not identify an intervention setting [77].

Twelve interventions were individual based, of which ten were reported as successful [21, 38, 48, 57, 63, 66, 70, 73, 78, 79] while two were unsuccessful [60, 80]. Seventeen interventions were group based, of which fourteen were successful [19, 41, 46, 52, 54, 58, 59, 61, 62, 64, 65, 72, 74, 76]. Ten interventions combined more than one method of delivery (e.g., group and individual), of which nine were successful [43, 45, 46, 49, 53, 56, 69, 71, 75]. The number of sessions ranged from 3—22, with large variations in frequency. Web-based interventions ranged in duration from 3—22 weeks.

A variety of educational approaches were used across interventions including lectures, counseling, informational booklets, tests, workshop discussions, interactive methods, motivational strategies and social support strategies. Multiple approaches were combined in many interventions performed face-to-face, digitally, or by telephone. A variety of resources were used to support intervention delivery including diary logs, reflective journaling, computer tablets, and developed learning modules. Details of health care professional (HCP) training were not generally reported.

In terms of content, 26 interventions reported a theory-based foundation. In total, 13 theoretical frameworks were used across studies in support of intervention content, revealing a broad conceptual understanding of empowerment. The most recurrent underlying theories were patient empowerment, presented in 10 studies [43, 45, 46, 48, 49, 59, 62, 64, 71, 72, 75]. Twelve studies did not use any theoretical framework, and four studies referred to theories to define concepts only without providing further evidence of use [52, 58, 70, 78]. The most common theory, the Person-Centered Model for the Promotion of his/her Empowerment [42] was used by four studies [41, 62,63,64]. Three studies [43, 45, 71] used the Health Empowerment Theory [44] and three studies [59, 72, 75] used the Family-Centered Empowerment Model [82]. Two studies [49, 69] used Educational Theory [50] and two studies [57, 65] used the Theory of Salutogenesis [83]. In addition, Almeida, Correira de Sousa, et al. [39] and Lenjawi, Mohamed, et al. [61] used the Health Belief Model [40]. Other theories such as the theory of problem solving [37] and the coaching framework [48] were represented by single studies.

Intervention content varied depending on study aims and assessed outcomes. Twelve studies aimed to understand the effectiveness of empowerment strategies on self-efficacy, self management behaviors, or readiness to make behavioral changes [19, 41, 43, 52, 56,57,58,59, 63,64,65, 71]. For example, Sit, Chair, et al. [71] evaluated a 13-week empowerment intervention on self-management behavior, self-efficacy and functional recovery, delivered across six weekly nurse facilitated group sessions and four weeks of telephone follow-up.

Ten studies aimed to increase empowerment of patients, using patient-reported outcome measures (PROMS) [38, 39, 46, 49, 62, 70, 75, 77,78,79]. For example, Vahedian-Azimi, Miller, et al. [75] described a cardiac rehabilitation program using the Family-Centered Empowerment Model delivered through 21 support group webinars to improve physical and mental health of post myocardial infarction (MI) patients [75].

Four studies described empowerment interventions that were aimed at improving quality of life [37, 51, 72, 80].

Eleven studies used empowerment interventions to improve glycated hemoglobin (HbA1c) along with other metabolic measures and patient-reported outcome measures (PROMS) [21, 43, 48, 49, 53, 54, 61, 62, 66, 68, 73].

There were large variations in study duration, follow-up time and measurement points. Most studies had a relatively short follow-up (2–12 weeks), but ten studies (26%) collected data after 6–9 months of follow-up.

Only five studies (12.8%) measured effects 1–2 years post-intervention [66, 68, 69, 75, 80].

WHO empowerment components

We assessed intervention content in relation the WHO conceptual framework of empowerment [15] by applying the four fundamental components and assessing whether they were incorporated within intervention design: (1) patient participation and understanding of their role; (2) patient acquisition of enough knowledge so they can engage with their health care provider; (3) patient skills; and (4) the creation of a facilitating environment [15] (Table 3). Patient skills (35/39) and patient participation (29/39) were addressed in most studies. Seven studies clearly described patient knowledge that enabled better engagement with HCP. All studies incorporated at least one component and 13 studies targeted three components. No studies addressed all four components. Components reflecting knowledge, making patients able to engage with the health care provider and the facilitating environment were scarcely reported.

Table 3 Overview of empowerment components in the included studies according to the WHO definition of empowerment
Full size table

Outcomes and instruments

Empowerment was presented as the primary outcome in 11/39 studies [37,38,39, 56, 58,59,60, 70, 76, 78, 79]. Clinical outcomes were presented as a primary outcome in 12 studies. In several studies, empowerment was used as a secondary outcome; in the remaining studies primary and secondary outcomes were not defined (Table 4).

Table 4 Overview of measurements used in the included studies of empowerment intervention
Full size table

The diverse array of outcome measures included: empowerment, self-management, sense of coherence, illness perception, anxiety and depression, self-efficacy, QOL, knowledge, self-care management, medication adherence, diabetic foot prevention, patient enablement, and post-traumatic growth. Clinical outcomes included: HbA1C; total cholesterol; triglycerides; high- and low-density lipoproteins; serum creatinine; and fasting/non-fasting blood sugars. Anthropometric measurements included waist circumference, body mass index, ejection fraction, and blood pressure.

The most common measurement instruments used included: the Diabetes Empowerment Scale-Short Form (DES-SF) alone or in combination with other instruments (n = 7) and the Diabetes Quality of Life (QOL) measures (n = 3) (Table 4).

Many studies combined several Patient Reported Outcome Measures (PROMS) or used these in combination with clinical measures. A total of seven specific empowerment-focused PROMS were used in 14 different studies. Different variations of the Diabetes Empowerment Scale were used in nine studies. Eight studies used self-management or self-care PROMS.

Intervention effects

Few studies were assessed as similar enough to be included in meta-analysis. Six studies reported group-format interventions using variants of the Empowerment Scale [46, 58, 77]; three of which also included an additional individual follow up, [49, 56, 68] with a total of n = 1,034 patients. Pooled results showed strong evidence for an effect favoring interventions, however with high heterogeneity (SMD 3.08; 95% CI, 1.95 to 4.22, p < 0.0001; I2 = 99%, p < 0.00001) (Fig. 2A). In six other studies of group-format interventions representing n = 1,434 participants, [45, 48, 49, 54, 61, 68] the pooled result showed strong evidence for a reduction in HbA1c (MD, − 0.32; 95% CI, − 0.47 to − 0.17; p < 0.0001; moderate heterogeneity I2 = 51%, p = 0.07) (Fig. 2B). Four other studies of group-format interventions [41, 52, 64, 69] also showed strong evidence of a pooled effect on self-efficacy (MD 1.86 95% CI, 0.81 to 3.24), p = 0.001; however, with high heterogeneity I2 = 99%, p < 0.06 (Fig. 2C). Four studies showed evidence of a positive pooled effect on self-management / self-care (MD 7.69), p =  < 0.001, I2 = 0% (Fig. 2D) [45, 49, 69, 71].

Fig. 2
figure 2

Forest plot of the meta-analysis of group-format empowerment interventions versus control using the Empowerment Scale (2A), HbA1c (2B), self-efficacy (2C) and Self-Management / self-care (2D)

Full size image

Three individual-format interventions measuring HbA1c found statistically significant improvement (MD, -0.33; 95% CI, − 0.59 to – 0.06; p = 0.02; (with high heterogeneity I2 = 87%, p = 0.0004), (Fig. 3) [21, 38, 48].

Fig. 3
figure 3

Forest plot of the meta-analysis of individual-format empowerment interventions versus control using HbA1c (E)

Full size image

Of all studies, 32/39 reported strong evidence for changes on the primary outcome of interest in favor of the intervention group. Of these, 10/32 were conducted in individual format; 15/32 were conducted in group format; and 6/32 in mixed formats.

One-third (13/39) of studies that included an empowerment measure found significant improvement in empowerment scores [38, 39, 46, 49, 56, 58, 62, 68, 70, 77,78,79,80]. Six studies also found improvement post-intervention in self-care management measures.

Of 12 studies that measured self-care management behavior, self-efficacy or readiness to make behavioral changes, effects were found in eight studies for example empowerment and foot care behavior, [56] empowerment and self-care behaviors, [58] and self-efficacy [43, 57]. Of ten studies that aimed to improve patient empowerment, eight were focused on type 2 diabetes, and all reported improvement in empowerment and self-care management [38, 39, 46, 49, 62, 70, 78, 79]. The four studies describing empowerment interventions aimed at improving QOL, optimism and control over life found mixed results, for example Tabari, Razi SH, et al. [72] reported an improvement in QOL among elderly people with COPD.

Quality appraisal

All studies were assessed using the Risk of Bias 2 (ROB2) tool, with 19% of the individual randomized studies evaluated as having an overall low risk of bias [84]. More than half of these studies (56%) showed high risk of bias, and a quarter had some concerns. In studies with a overall high risk of bias, concerns arose primarily from the randomization process and/or possible deviation from the intended intervention. However, in terms of selective reporting of results, more than three-quarters of studies (30/39) received a low ROB score (Figs. 4 and 5).

Fig. 4
figure 4

Risk of Bias Domains for individual randomized studies using Rob2 tool. Domains: D1: bias arising from the randomization process, D2: bias due to deviations from intended intervention, D3: bias due to missing outcome data, D4: bias in the measurement of the outcome, and D5: bias in the selection of the reported result. Legend: Red (x) = high risk of bias; Yellow (-) = unknown risk of bias; Green ( +) = low risk of bias

Full size image
Fig. 5
figure 5

Risk of Bias Domains for cluster randomized studies using Rob2 tool. Domains: D1: bias arising from the randomization process, D1b: Bias arising from the timing of identification and recruitment of individual participants in relation to the timing of randomization. D2: bias due to deviations from the intended intervention, D3: bias due to missing outcome data, D4: bias in the measurement of the outcome, and D5: bias in the selection of the reported result. Red (x) = high risk of bias; Yellow (-) = unknown risk of bias; Green ( +) = low risk of bias

Full size image

Among the seven cluster randomized studies, only one was evaluated as having a overall low risk of bias [66]. All studies scored low risk in bias due to missing outcome and bias in selection of the reported results. The most problematic domains in these studies were bias due to deviations from the intended intervention and bias in measurement of the outcome. Four studies were evaluated as having some concerns in overall bias and two studies had a high risk of bias overall (see Supplementary file 2 (Table 5) for reasons for the selected assessments and Supplementary file 3 for a Graph Summary plot (ROB2).

No studies were excluded from the review or effect presentations due to poor methodological quality.

Discussion

This review of empowerment interventions covering thirteen diagnostic categories expands on prior findings regarding content, measures and efficacy of empowerment approaches in chronic disease.

We found that a majority (58%) of studies used a theory or framework, a finding that contrasted with Werbrouk et al. [3] who detected a much larger proportion of studies (81%) that employed a theory-based intervention. These findings suggest that incorporation of theory in intervention design has declined in recent years. We also reviewed intervention setting, modality and content to clarify consistency with WHO’s empowerment definition. Overall, we found little consistency in how empowerment was deployed conceptually in the design of interventions and we did not identify any studies that included all four of WHO’s fundamental constructs for empowerment. Most studies (85%) employed just two components. However, WHO describes these components as fundamental, signifying that each is equally important for establishing efficacious interventions. This approach is supported in assessments of the concept within the literature that consider empowerment as a dynamic process that addresses the sense of powerlessness and loss of control that is common among individuals’ who are a managing a chronic disease [5]. Aujoulat et al. [85] described empowerment as consisting of an inter-personal dimension (a process of communication and education in which knowledge, values and power are shared in provider-patient interactions) and an intrapersonal dimension (patients’ process of personal transformation). Dialogue between health care providers and patients, co-creation of knowledge, a patient-centered approach, enhancement of patient competencies, and active participation have also been identified as antecedents of patient empowerment while self-management and improved quality of life have been identified as potential outcomes of the empowerment process [11, 86]. The WHO definition is generic and not specific enough on the above mentioned aspects of the empowerment process, however, the first fundamental component includes the concepts of patient participation, patient knowledge and patient skills, and therefore, it reflects a person-centered perspective to a certain degree. Based on our study and use of the WHO, we recommend the development of an updated and unified definition of empowerment that capture the importance of the person-centered perspective and emphasize the dialogue with health care professionals in order for empowerment to happen. There is a need for a more thorough analysis of the personal transformation which develops the individual`s ability to cope and the transfer of power between health care professionals and patients. Furthermore, Health coaching has developed as an important approach to promote self-awareness and empowerment in patients with chronic disease [87,88,89] and is an interesting field for future studies.

Notably, nurses were responsible for delivering the intervention in half (56%) of studies. In addition, some studies described interprofessional collaboration in intervention delivery where nurses often were team-members. These findings are consistent with the philosophy of patient-centered care, which implies patient activation and patient participation in practice [90]. According to WHO [15], patient participation is the first of the fundamental components of empowerment. In order to utilize patients’ inherent resources for health, tools and interventions to exploit these under-utilized resources are needed [10]. Patient participation and patient activation can therefore be seen as complementary strategies for achieving patient-centered care, which in turn can affect patient empowerment [11]. Most studies were implemented in outpatient or community-based settings where community and public health nurses often have the responsibility for intervention delivery and follow-up of persons with chronic disease. Primary care is a highly relevant setting for the development and delivery of empowerment-focused strategies and interventions by interprofessional teams and by nurses independently.

In descriptive analyses, we found that most interventions delivered in group-format (13/17) and individual-format (10/12) were reported as successful. Of ten interventions that combined more than one method of delivery (e.g., group and individual), nine were successful. These findings diverged from the meta-analysis undertaken by Werbrouk et al. [3] of 23 empowerment interventions among patients with somatic chronic diseases that found an effect in favor of individual format interventions on empowerment-based PROMS. In contrast, our results suggested that empowerment interventions were more effective when conducted in groups or in combinations of group and individual formats. In our meta-analyses of six studies in group-formats measuring HbA1c, we also found strong evidence in favor of empowerment interventions, which did not hold in individual formats. These findings were consistent with a meta-analyses of 21 studies that compared group-based diabetes self-management education with routine treatment, a waiting list control and no intervention, finding strong evidence for an effect on HbA1c at 6, 12, and 24 months of follow-up [91]. The superiority of the group-directed format of empowerment interventions was also confirmed in Chen et al.’s review [23] that found improvements in blood pressure and reductions in cholesterol among people with diabetes.

Our review also found that empowerment interventions were effective when measuring several clinical markers and PROMS, including empowerment, self-efficacy and self-care management. These findings aligns with results reported by Chen and I-Chuan [23] suggesting that empowerment interventions improved the health status, psychological status and quality of life among patients with chronic disease. These findings also confirm that development of broadly applicable empowerment interventions may be a promising approach for future intervention development focused on improving self-care management and health among patients with chronic disease.

We summarized outcome measures used to evaluate empowerment, self-care management and clinical outcomes and found inconsistencies in measurement. According to WHO, empowerment is a unique concept with the potential to influence patient activation and self-management [15]. However, we found that only 36% of studies used any empowerment scale, 31% used clinical outcome measures, 21% measured self-management, but only 5% of studies used all three measures. These findings reveal a gap in knowledge on the essential role of empowerment in pathways that include patient activation, self-care management, and clinical outcomes.

Recent disease specific reviews had been conducted among patients with hypertension [92], chronic metabolic diseases [26] and people with type 2 diabetes mellitus [91]. Our review contributes to highlighting the importance of empowerment interventions among patients with chronic disease in general. We assessed the overall quality of the evidence and found that only one in five studies had a low risk of bias while the majority were high risk. Other recent reviews found similar levels of quality. In their review of internet-based intervention studies, Kuo et al. [26], found that fewer than one-third (29%) of the 21 reviewed studies reported allocation concealment, blinding of outcome assessments, or role of study personnel. Overall, our review revealed inconsistencies concerning definition of patient empowerment consistent with Mora et al.’s [12] descriptive review. Using the WHO [15] definition of the empowerment process and ensuring that all four fundamental components of empowerment are covered in intervention design, may provide more consistency in future research and clinical practice. Future studies on patient empowerment should also consider including both an empowerment measurement tool as well as measures of self-management and clinical outcomes to assess the effect of empowerment strategies.

Strengths and limitations

Strengths of this review included the wide range of settings and populations included and consistent outcome measures that enabled meta-analyses on individual and group format of the interventions as well as on some PROMS. To our knowledge this is the first review of empowerment interventions including such a diversity of chronic diseases. Limitations included the possibility that we could have missed reports not indexed within the six databases searched, from references cited within our included studies, and in grey literature. Another limitation is that eHealth studies has not been included if they did not include empowerment in the title or abstract. There are many important concepts related to methods and approaches in empowerment interventions that could have been used in the search, i.e. patient participation, patient activation, patient engagement, shared decision making, health coaching and more. Our choices may have had the consequences that we have lost some studies that otherwise might have added to the findings. The sample size of most studies included in our meta-analyses was small. The intervention effect on the PROM measures should be interpreted with caution due few studies eligible for inclusion and high heterogeneity of modes, operational definition of empowerment and measurement tools. Furthermore, it was challenging to extract and categorize interventions because of considerable variability in intervention design.

Conclusion

In conclusion, our findings demonstrate that empowerment interventions in chronic disease contains essential components that contribute to strengthening patients’ capability for self-care management and health in chronic disease and are important in order to attain WHO`s sustainable development goals. Future studies investigating the role of empowerment in chronic disease should consolidate conceptual understandings by using WHO’s empowerment components and investigate the role of empowerment in pathways that include patient activation, self-care management, and clinical outcomes. Group-format or mixed format interventions delivered in outpatient or community health settings and Primary Care are especially suitable to facilitate patients’ process of taking control of their health.