Feasibility of acquiring occupational data from cancer patients’ medical records in Malaysian oncology centres | BMC Public Health

Work related cancer is not easily recognized as it can appear years following initial exposure or after retirement. Identifying the link could be made more difficult among those with multiple occupations. Thus, a detailed occupational history, which includes previous work experiences and work exposures are important in identifying probable work-related cancers [2, 4]. This study found limited information on occupational details reported in the medical records from both oncology centres. The available data were also inadequate, mostly only limited to work classification (81.1%) with almost no information on exposure at the workplace (0.2%).

The early screening process showed that the centre with paper-based record keeping system, where occupational history was written manually, had a higher percentage (88.5%) for availability of occupational information compared to centre with EMR (85.3%). The EMR system used in the studied facility includes a designated free-text field for occupational information, although it did not enforce mandatory completion. As such, the recording of occupational information depends heavily on physician awareness and initiative. Additionally, limited access to computers and lack of point-of-care documentation have been shown to delay data entry in EMR leading to data inaccuracy [18]. This could also be influenced by the complexity and skills required to use the system efficiently and the lack of prompt for the entry of occupational information [19, 20]. Other studies have also reported dissatisfaction among clinicians with the completeness and correctness of data recorded in EMR [21, 22].Additionally, the system can be more time consuming to navigate compared to handwritten records, leading to reduced time for patient interactions [23, 24]. This limitation also reflects the broader institutional or higher-level decisions made during procurement of the systems to be used in the facility.

Despite the lower availability of occupational information in EMR from the screening data in this study, more information on specific type of occupation details were able to be categorized from this system. Higher availability of specific occupational details including job title, work classification, work sector, workplace and exposure were observed from EMR. This contrasted with the observation from one study that reported higher chance of data entry errors in the free-text field, potentially contributing to incomplete occupational information [20]. Given the circumstances, with mandatory requirements or system design to capture detailed occupational information including exposure history in the EMR, more complete and relevant data could be documented.

This study showed a higher availability of occupational data compared to a study in North Carolina which had about 40.3–42.2% data for type of occupation, and 43.5–42.2% for type of industry being reported from 2020 to 2021 [25]. Another study analysed occupational data in the Florida cancer registry among firefighters. They reported approximately 53% of occupational code were available to enable classification of workers into detailed occupations [26]. Meanwhile, data on the duration of work, hazardous exposure and previous work were severely lacking with less than 2% of records having this information. Of note, the lack of exposure data imposes a great challenge to physician to attribute the cancer causes to specific exposures or risks [27]. Based on a review done on the occupational cancer burden in Great Britain, the population attributable fraction of total cancer as a result of occupational exposure ranged between 2% and 8% [28]. Whereas, in Italy, 0.9% of cancer diagnosed and 1.6% of cancer mortality were attributable to occupational exposure to carcinogens [29]. This implies the importance of exposure data, as a substantial proportion of cancer diagnoses were due to occupational exposure.

When compared with patients’ interview in this study, only 65.3% showed similarity in work classification as documented in the medical records, whereas, 8.9% showed total discrepancy. The discrepancy was higher compared to a study in Norway which reported 5% inconsistent information between electronic and paper-based record [30]. The discrepancy could also be attributed by previous occupational histories that were not captured in either the record or by the interviewer. Additionally, using systematic and comprehensive checklists that encompasses key occupational information during history taking can help in preventing discrepancies and missing data. Studies have shown that systematic screening of occupational exposures can improve the identification of work-related cancers and support compensation efforts [31,32,33].

While history taking is a key component in reaching a diagnosis, occupational history is often neglected, leading to underdiagnoses of work-related cancers [14]. Occupational information is often perceived as neither useful nor urgent, and the assessment is viewed as complex [34]. The lack of knowledge and awareness among medical practitioner on the importance of occupational information and the details required to establish work-related diseases, could influence the quality of the data collected [35]. Specifically for exposure, the intensity and duration are important information needed to establish the work-relatedness of the cancer outcome [32]. Additionally, the lack of standardized definitions of exposure and diagnostic criteria for occupational diseases impedes the process of collecting this information [36]. A comprehensive occupational history is therefore important in establishing a link between occupation and disease, which subsequently can expedite work-related diagnosis, prompt treatment, and facilitates compensation process [14].

In this study, data on occupations are mostly available in medical records. However, the depths of information are scarce. Most information is limited to only the current or last job title, with little information on previous jobs, the work activity, exposure at workplace, intensity or duration of exposure that could provide information to possible causes of work-related cancer.

Strengths and limitations

To the authors’ knowledge, this study was the first to investigate the acquisition of occupational data from cancer patients’ medical records in Malaysia. This pioneering effort is crucial in assessing how well attending medical practitioners documented occupational histories. The study provided valuable baseline data on the availability of occupational history, highlighting gaps in collection of information. Understanding these gaps is essential for improving documentation on occupation during patient consultations. Comprehensive occupational details could establish a link between occupation and cancer, which is important for diagnosis, planning and implementation of preventive strategies, and compensation claims.

However, this was a descriptive feasibility study with a small sample size involving two centres in Malaysia. Furthermore, the patient interviews were conducted only at a single site, which was the paper-based oncology centre. Although the findings from the interview component may not be generalizable to other settings and may introduce potential bias, it provided valuable insight into the discrepancies between patient-reported occupational history and the information recorded in medical records, highlighting specific gaps in documentation practices. In addition to that, recall bias may have affected the accuracy of self-reported occupational information, particularly among patients with long work histories or complex employment backgrounds. Moreover, present study did not explore the underlying reasons for gaps in documentation of occupational data or discrepancies between patient interviews and medical records. Hence, additional studies are needed to investigate the reason behind incomplete or inconsistent occupational information in records. Specifically, research focusing on medical practitioners’ practices and factors contributing to these discrepancies would provide deeper insights into improving data collection methods.