Data sources
This is a retrospective cohort study using a national database, the Taiwan National Health Insurance Research Database (NHIRD). The NHIRD contains registration files and original claims data including International Statistical Classification of Diseases (ICD) codes for reimbursement for approximately the entire population of Taiwan [10]. After data de-identification, the NHIRD is provided to the scientists in Taiwan for research purposes [10].
Identification of dentists, general population, and other HCPs
We identified all the dentists from the 2009 registry for medical personnel in the NHIRD (Fig. 1) [11,12,13,14]. The 2009 registry for medical personnel we used included all the HCPs registered in 2009 and their medical histories between 1996 and 2012. An identical number of age- and sex-matched participants from the general population was identified for comparison. In order to decrease the number of potential confounders, we also identified a cohort consisting of other HCPs who may share similar working environments and socioeconomic status with the dentists for comparison [14,15,16]. The other HCPs included all the physicians, pharmacists, medical technicians, audiologists, consultant experts, clinical experts, dietitians, social workers, and language experts identified in the database we used, which were not matched with the dentists or general population. Age subgroups were categorized as ≤34, 35–59, and ≥ 60 years [14,15,16,17]. The underlying comorbidities were defined as hypertension (HTN) (ICD-9-CM code: 401–405), hyperlipidemia (ICD-9-CM code: 272), liver disease (ICD-9-CM code: 570–576), mental disorders (ICD-9-CM code: 290–319), diabetes mellitus (DM) (ICD-9-CM code 250), coronary artery disease (CAD) (ICD-9-CM code: 410–414), chronic obstructive pulmonary disease (COPD) (ICD-9-CM code: 496), malignancy (ICD-9-CM code: 140–208), stroke (ICD-9-CM code: 436–438), and renal disease (ICD-9-CM code: 580–593) [14]. The underlying comorbidities included in this study were based on having received the diagnosis during at least three ambulatory care visits, or at least one hospitalization [14, 15]. We included and adjusted for these underlying comorbidities in the logistic regression to control confounding effects [14, 18]. The participants who had been diagnosed with C-HIVD (ICD-9-CM codes: 722.0, 722.4, 722.71, or 722.91) before 2007 were excluded. C-HIVD was defined as the participants who received the diagnosis during at least one hospitalization or ambulatory care. We excluded the participants with C-HIVD before 2007 because we wanted to calculate the cumulative incidence of C-HIVD by following up between 2007 and 2011.
Comparison of the risk of developing C-HIVD
We followed up the development of C-HIVD in the participants between 2007 and 2011 to compare the cumulative incidence of C-HIVD between dentists and general population, and between dentists and other HCPs. To assess whether age and sex were effect modifiers, stratified analyses for age and sex subgroups were also performed.
Ethics statement
This study was approved by the Institutional Review Board at the Chi-Mei Medical Center. We conducted this study strictly according to the Declaration of Helsinki. This is a secondary data analysis article from the NHIRD and all necessary permissions were obtained to access and use the data. Informed consents from the participants are waived because the NHIRD contains de-identified information, which does not affect the rights and welfare of the participants.
Statistical analysis
An independent t-test for continuous variables and a chi-square test for categorical variables were used in the comparison of demographic characteristics and underlying comorbidities between dentists and members of the general population. Conditional logistic regression analysis was performed to compare the risk of developing C-HIVD between dentists and general population by adjusting for HTN, hyperlipidemia, liver disease, mental disorders, DM, CAD, COPD, malignancy, stroke, and renal disease. Unconditional logistic regression was performed to compare the risk of developing C-HIVD between dentists and other HCPs by adjusting for age, sex, HTN, hyperlipidemia, liver disease, mental disorder, DM, CAD, COPD, malignancy, stroke, and renal disease. The 5-year (i.e., follow-up period between 2007 and 2011) cumulative incidences of C-HIVD in the dentists, general population, and other HCPs were also calculated. We used SAS 9.4 for Windows (SAS Institute, Cary, NC, USA) for all analyses. The significance level was set at 0.05 (two-tails).