The electronic searches returned 3825 records; of those, 2930 unique records were screened after removing duplicates (Fig. 1). Following screening, 50 studies were included, comprising 20 clinical trials and 30 observational studies.

PRISMA diagram. PRISMA Preferred Reporting Items for Systematic Reviews and Meta-analyses, RCT randomised controlled trial

Figure 2 describes the study design of the included studies, and Table 2 summarises their characteristics. Table S2 provides additional details of study characteristics. Half of the clinical trials (10/20; 50%) applied a randomised parallel design, while nearly one-third of observational studies (9/30; 30%) were case series or case reports. Over half of the clinical trials were conducted in Asia (55%; 11/20) [15,16,17,18,19,20,21,22,23,24,25], while most observational studies took place in European (43%; 13/30) or Asian countries (37%; 11/30) [26,27,28,29,30,31,32,33,34,35,36,37,38]. Study settings at the time of patient enrolment were not reported in 36% (18/50) of the included studies. Most studies reporting this information recruited participants through outpatient settings, which accounted for 58% (7/12) of the clinical trials [17, 18, 22, 39,40,41,42] and 65% (13/20) of the observational studies [27, 29, 32, 34, 35, 43,44,45,46,47,48,49,50]. Information on the study period of enrolment was available from 54% (27/50) of the studies included in this review. Among the clinical trials that provided information about their enrolment period, most (9/10; 90%) enrolled patients from year 2000 onwards [15, 17, 20,21,22, 25, 39, 41, 51]. One clinical trial was conducted from 1987 to 1991 (totalling 4 years of enrolment) [16]. Among the observational studies that reported the enrolment period, 71% (12/17) enrolled patients from 2000 to 2018 [10, 28, 34, 38, 43, 46, 47, 49, 52,53,54,55]. The remaining studies (5/17; 29%) were conducted between 1976 and 1988 [27, 35, 36, 48, 50].

Study design of the included studies. RCT randomised controlled trial

The studies included in this review documented HPL among patients receiving various drug treatments. However, none were designed with the primary objective of assessing the causal relationship between the drug treatment, the development of HPL, and the symptoms experienced by patients. Instead, patients were identified with drug-induced HPL through different clinical approaches describing the drug treatment as the primary suspected cause of the HPL. Among the included studies, the definition of HPL varied substantially. In general, PRL levels used to define HPL ranged from ≥ 10 ng/mL to ≥ 30 ng/mL. Some studies defined specific PRL levels for males and females to diagnose HPL, while others applied different levels of PRL excess to classify mild and severe HPL cases. Study-specific definitions of HPL are provided in Table S2.

Finally, the duration of treatments suspected to have caused HPL differed among studies, and half of the included studies (25/50; 50%) did not report data on treatment duration. Fifteen clinical trials (of 20; 75%) reported a study duration ranging from 6 weeks to 2 years, and most trials (10/15; 67%) lasted between 2 and 4 months [17, 18, 21,22,23,24, 39, 42, 56, 57]. Ten observational studies (of 30; 10%) reported study durations varying from 2 months to 24 months. Most observational studies (7/10; 70%) lasted 6 months or less [26, 29, 44, 45, 47, 49, 58, 59].

Table 3 summarises the demographic characteristics of patients in the included studies. Among the clinical trials and observational studies, samples comprised mostly adults (42/50; 84%). One clinical trial [41] and two observational studies included children [28, 49], and one clinical trial [51] and four observational studies enrolled both children and adults [26, 34, 52, 53]. Most studies recruited both females and males; however, females were slightly more prevalent among patients with drug-induced HPL. Among the clinical trials, half of the studies (10/20; 50%) assessed at least one female-only treatment group [15,16,17,18,19,20,21,22, 25, 60], while the proportion of females ranged from 0 to 90% in the remaining trials [23, 24, 39,40,41,42, 51, 56, 57, 61]. Similarly, 77% (23/30) of the observational studies reported a proportion of females as ≥ 50%, with 37% (11/30) reporting data on females only for at least one of the treatment groups assessed.

Two-thirds (33/50; 66%) of the included studies reported on the serum PRL level of patients with drug-induced HPL. Mean PRL levels of patients with drug-induced HPL varied from 32 to 168 ng/mL [17, 24]; one trial reported median and range values (84 ng/mL; 44–99 ng/mL) [42], and two studies reported only ranges [40, 60]. Over two-thirds of studies reported serum PRL levels (19/30; 63%); of those, 21% (4/19) described data among the overall set of participants enrolled in the study, regardless of the drug-induced HPL status [32, 34, 54, 55]. Fifteen studies (of 19; 79%) reported data on patients diagnosed with drug-induced HPL and assessed in at least one treatment group. Mean PRL levels of patients with drug-induced HPL ranged from 24 to 145 ng/mL [44, 46, 48]. Two studies reported only ranges [