The reference lists in the selected articles were screened, and 125 references that could be of interest were imported to Covidence, and the procedure was followed. Altogether 1914 abstracts were screened (Fig. 1).

Flow chart for search strategy and selection of studies

Eighty-four full-text articles were assessed, of which 54 were excluded as they did not fulfil the criteria (Fig. 1). From the remaining 30 articles, an extraction of relevant information was made by two blinded authors in random pairs. Apart from first author, title, and year of publication the following information was extracted: type of study (qualitative, cross-sectional, longitudinal or review); country where the study was performed; disorder (type of injury/disease); type of work or exposure and receiver of communication (employer, employee or group of employees). In addition, we also interpreted, assessed and categorised the articles based on aspects of communication [results concerning an individual or aggregated group results, the sender, the timing, the location, the content, the mode (i.e., written, oral, face to face, telephone or other), the receiver, or other aspects]. More than one aspect could be indicated. Further, we extracted relevant results, conclusions, and key points, as well as other relevant comments from the studies. Finally, all three authors compared these extractions and decided on which information to report.

The final publications selected included 30 articles published between 1980 and 2023 [9, 14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42]. Their descriptive characteristics are presented in Table 1. None of them studied feedback after health surveillance targeting hand-arm vibration injury, or even surveillance that was performed due to hand-arm vibration exposure. Still, we considered the results and the authors’ reflections valuable for workers’ health surveillance and therefore chose to present them.

Ten of the included studies were carried out in the Netherlands and six in Sweden, followed by the USA with five studies and the United Kingdom with four (Table 1). Two studies were conducted in Italy. Australia, Belgium, and Israel were represented with one study each.

The majority (n = 16) of the included studies were of longitudinal design, followed by qualitative studies (n = 7; Table 1). Four studies were cross-sectional, and three publications were reviews. The studies reported on workers’ health surveillance that addressed musculoskeletal disorders and pain (n = 8), risk of cardiovascular disorders (n = 4) or hearing disorder (n = 3), work ability and fitness for duty (n = 3), mental health (n = 2), allergy/ asthma (n = 2), and cancer (n = 1). There were additionally seven studies that addressed a mixture of disorders and general health (n = 7).

All the 30 selected studies addressed the employee as the receiver of the feedback of the health surveillance [9, 14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42].

Sixteen studies focused on factors related primarily to the sender [9, 14,15,16,17,18,19,20,21, 26, 28, 31, 33, 35, 38, 41]. The suggested feedback provider to an individual worker depended on the disorder at focus in each specific health surveillance. For example, in studies addressing musculoskeletal problems, physiotherapists and ergonomists were the primary information providers. For discrete diseases other than musculoskeletal problems, the specific disorder guides medical personnel (physicians, nurses, or audiologists) who possesses the expertise and could be an appropriate provider of the feedback.

The only question discussed concerning the sender was related to the medical legitimacy of the feedback provider. Should it be practitioners, occupational health physicians, or other medical specialists who preferably give feedback [16, 17]? In line with the professional backgrounds of the first authors of the various papers, ergonomists, physiotherapists, occupational health physicians, psychologists, audiologists, and nurses were suggested as feedback providers.

Di Battista et al. [17] highlighted the positive value of conveying trust, confidence, and a high level of knowledge and competence in the sender. A lack of confidence in the sender was observed when the risk assessment at the occupational health centre and the advice about medication from the general practitioners were contradictory. The study by Battista et al. is the only study specifically addressing personal aspects of the sender, as it highlights the importance of the sender being attentive to their own mindset and emotional state to minimize bias.

Several papers stressed a general need for the feedback provider’s competence. In the study by Boschman et al. [15], the occupational physicians reported that they needed even more knowledge on some of the preventive advice to counsel the recipient better, and that indicated that they needed more detailed training. A similar conclusion on the importance of competence of the performer of health surveillance for proper feedback was drawn by Codling et al. [16].

Menckel et al. [9] reported that the physiotherapists who delivered the feedback in their study on musculoskeletal disorders enjoyed the workers’ confidence, which was important for generating changed behaviours for prevention. Steel et al. [38] could, on the other hand, show that lower trust in the physician lowers the intention to disclose problems. Ruitenburg et al. [35] reported that when the sender was in the occupational health service, and the receiver of the feedback was a hospital physician, the latter presented more doubts. This might, however, concern the effectiveness of occupational health surveillance or a fear of medicalization among the hospital physicians.

There were altogether few studies that analysed the impact of factors specifically related to the sender of feedback and how this influences an effective perception of the feedback information.

Thirteen studies reported aspects on the importance of timing on when to give feedback [9, 16,17,18, 24, 26, 27, 29, 32, 33, 37, 41, 42] however, not considering that the impact of timing and choice of location are closely interrelated.

The severity of the outcome disease is suggested to determine the timing of when feedback should be given. In a screening program for bladder cancer conducted by Marsh et al. [32] they established that in the case of a positive screening result, an immediate diagnostic evaluation was warranted, together with an immediate telephone call and a written follow-up letter, while a non-negative result could be followed up within one month.

Health surveillance results providing information that causes mild anxiety has been reported as a motivating factor for behaviour change [26]. The authors suggest that when individuals receive information about their personal risk that may cause anxiety, advice and support should also be given, at the same time. This is with the pronounced intention of reducing anxiety. Other authors suggest that mild anxiety must not be avoided as fear and perceived threat might be a favourable factor for preventive change. Information received may, in addition to anxiety, for the recipient also lead to an overestimation of the risk status, e.g., concerning cardiovascular risk [37].

When risk assessment for cardiovascular disease, included e.g., HbA1c and cholesterol levels, instant feedback was given and reported. Immediate feedback was reported to facilitate engagement for prevention [17]. Urgent abnormal results were immediately notified by telephone [