Seventeen studies were included [29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45]. The study selection is illustrated in Fig. 1.

Flow of study selection process

A summary of the included studies is provided in Table 2. The majority of the included studies (n = 13, 77%) were published within the past five years. Eleven of the selected studies were RCTs [29, 31,32,33, 35, 37,38,39, 42, 45, 46] and six were non-randomized intervention trials [30, 34, 36, 40, 43, 44].

The total number of participants was 827, however, the studies varied considerably with regard to the number and characteristics of participants included. The median sample size was 34 participants per study. Peel and colleagues (2016) included the highest number of participants (n = 270). The majority of the studies (n = 14, 82%) included participants with a specific clinical diagnosis [29,30,31, 33, 35, 36, 38,39,40, 42,43,44,45,46]. The trials included patients with osteoarthritis [29, 39, 40], obesity [35, 38], chronic obstructive pulmonary disease (COPD) [33, 43], Morbus Parkinson [30], chronic kidney disease [31], chronic heart failure [36], kidney transplant recipients [41], and mild cognitive impairment [44]. The participants of two studies were cancer survivors [42, 45]. Three studies did not focus on a specific indication but included patients with various medical conditions [32, 34, 37]. In eleven studies, a home-based intervention in community-dwelling patients was implemented [30, 32, 33, 36, 38,39,40,41,42, 44, 45]. In two studies, the PA monitor-based intervention was conducted in an outpatient setting [31, 34] and three studied conducted the intervention during inpatient treatment [29, 37, 43]. The setting was not clearly specified in one study [35].

The intervention components used in each of the included studies are presented in Table 3. Four studies included two relevant intervention arms that were included as separate interventions [31, 34, 41, 42, 45]. Hence, a total of 22 interventions were included in the analysis. Table 4 shows the frequency of intervention components used in the included studies.

In nine interventions a body-worn PA monitor and BCTs facilitated by the PA monitor (i.e., feedback, goalsetting, self-monitoring) were used as main intervention strategy [31, 32, 34, 39, 41, 42, 44, 45], whereas in 13 interventions the PA monitor-based component was embedded into usual care, an indication-specific intervention and/or combined it with a structured behavioral intervention [29, 30, 33,34,35,36,37,38, 40,41,42,43, 45]. Nine interventions used a pedometer [29,30,31,32,33