Physician Suicide in Taiwan: A Nationwide Retrospective Study from 2000-2013

Research Article

Austin J Psychiatry Behav Sci. 2014;1(2): 1008.

Physician Suicide in Taiwan: A Nationwide Retrospective Study from 2000-2013

I-Ming Chen1, Shih-Cheng Liao1,2 and Ming-Been Lee1,2,3*

1Department of Psychiatry, National Taiwan University Hospital, Taiwan

2Taiwan Suicide Prevention Center, Taiwanese Society of Suicidology, Taiwan

3Department of Psychiatry and Social Medicine, National Taiwan University College of Medicine, Taiwan

*Corresponding author: Ming-Been Lee, Department of Psychiatry, National Taiwan University Hospital, No 7, Chung San South Road, Taipei, Taiwan

Received: February 10, 2014; Accepted: March 07, 2014; Published: March 10, 2014


Occupational health in medical professionals is an emerging issue in Taiwan; however, little research has specifically focused on suicide among doctors. In this study, we examined 1780 death records in a nationwide insurance registry covering all deceased physicians from 2000–2013, defined the probability of suicide for each record, and investigated the associated risk factors. Univariate logistic regression analysis showed that being male, born in China, and having a non–medical school background (special exam certified) were associated with a lower risk of suicide. A younger age at death, being born in countries other than Taiwan or China, and specialties including surgery, pathology or medical imaging, emergency, psychiatry and anesthesiology, increased the risk of suicide. After adjusting for confounders, only a young age at death was associated with a significant risk of suicide. The proportionate mortality ratios for doctors aged 25–44 years were 2.18 for males and 1.88 for females. Gas poisoning, hanging, and jumping were the most common forms of suicide among young physicians, while poisoning was most common in middle age. Our findings indicate that physicians may be more vulnerable in their early career years. Supportive interventions focusing on young physicians may promote the mental health of doctors.


Suicide has a strong link to occupation, and an increased suicide rate may reflect particular risk factors imposed on specific professional groups [1]. Physicians constantly have to bear heavy responsibilities, long working hours, emotional strain, service system reforms, and lawsuits, and are therefore more vulnerable [2–5]. On the other hand, medical practitioners are devoted to enhancing the health of their patients, are well–trained in medical knowledge, have advantageous socioeconomic resources, and may live longer than others [6,7]. Therefore, whether or not physicians are at a higher risk of suicide remains a topic of interest.

The nature of suicide is distinct from other causes of mortality among physicians. A previous study concluded that physicians have a lower mortality rate from all causes of death compared with the general population, except for suicide [8]. Another study also stated that physicians were more likely to die from stroke, accidents, and suicide, while the rates of other causes of mortality were lower [6]. In many countries, researchers have reported that the risk of suicide among physicians is higher than that in the general population and for other professionals [6,9–11]. It has been reported that the estimated suicide rate of physicians ranges from 19 to 69 per 100,000 [9–12], and that this risk is 2– to 4–fold higher than that of the general population [10,11,13]. These findings suggest that death by suicide involves additional risk factors than other causes of death. However, Shang et al. reported that the overall standardized mortality ratios (SMR), including suicide (SMR = 0.14), were lower in physicians than in the general population in Taiwan, which is in contrast to studies from other countries [7].

Many studies have investigated the factors associated with physician suicide. Some have found that the physicians who commit suicide suffer from depressive symptoms and other psychopathology, or had inadequately treated mental illnesses [14–19]. Other studies have found that job unsatisfaction, stress, and burnout can also lead to physician suicide [3–5,15,18]. Environmental exposure or access to abusive or lethal drugs [20], being female, living alone [16], particular personality traits [21–23], and certain specialties [12,24] are also considered to be potential factors contributing to physician suicide. In addition, cultural differences in attitudes toward suicide may influence the prevalence of suicidal ideation among medical students, and consequently the extent of actual attempts [25].

Although a previous study revealed that doctors in Taiwan have a lower suicide SMR than the general public, generalization of this result was limited by the confounding effects of educational background and socio–economic status [7]. In addition, the SMR may not represent the special characteristics of suicide concealed in the overall health–promoting nature of the medical profession. The insurance database of the Taiwan Medical Association includes records of all deceased physicians in Taiwan. Pan et al. explored physician suicide by descriptive analysis using this registry database in a preliminary study [26]. In the current study, we demonstrate the distinct impact of suicide among all–cause mortality, and further analyzed the associated risk factors for suicide for doctors in Taiwan.


All practicing doctors in Taiwan are required to enroll in death insurance managed by the Taiwan Medical Association. Whenever a physician dies, the personal registry in the national database managed by the Taiwan Medical Association is updated. The original causes of death listed on death certificates issued by treating doctors or coroners are recorded in the registry that families can receive death benefits. The available data from Taiwan Medical Association were delinked from the names and personal identification numbers of the deceased physicians.

We retrieved all 1780 records registered in this database from January 1, 2000 to May 9, 2013. In addition to the dates and causes of death, data including the dates and place of birth, gender, medical school from which the physician graduated, specialties, year of license registration, serving institutions and their locations were also collected.

The data were scrutinized by the first author who is a boardcertificated psychiatrist with an additional background in internal medicine residency. All authors confirmed the causes of death in a consensus meeting. The ages at death were grouped into 25–44, 45– 64, and 65 years or more. Places of birth were grouped into: Taiwan, China, and other foreign countries. In the early 1970s, Taiwanese government held special qualification examinations for doctors who have not attend regular medical college education but were trained in the military, and issued medical license for those who were qualified. Therefore, we categorized educational background into domestic medical school, foreign medical school, and nonmedical school (special exam certification). The specialties were classified into general⁄internal medicine (including general medicine, family medicine, physical medicine, neurology and pediatrics), surgery (including dermatology, otolaryngology, ophthalmology, obstetrics, gynecology, orthopedics), pathology or medical imaging (nuclear medicine, radiotherapy, radiation oncology, and radiology), emergency, psychiatry, and anesthesiology. The practice settings, reviewed by the name of hospital, were categorized into private clinic, district⁄regional hospital, medical center, or retired (without mention of previous service settings).

The original data regarding the causes of death were recorded in a primitive and unprocessed form. For example: “pneumonia; septic shock; end–stage lung cancer with brain metastasis”, “suicide with drug overdose (anesthetics)”, “falling in the bathroom”, or “suffocation; burning charcoal in the room”. We evaluated the possibility of suicide in each death for each record in this database. Accordingly, each record was assigned to one of the three defined extents of suicide: definite, probable, and possible suicide. Definite suicide was defined as deaths clearly specified as suicide, such as “suicide”, “hanging”, “jumping from a height”, and “poisoned by succinylcholine and midazolam”. Probable cases were identified from obscure descriptions such as “drowning”, “suffocation”, “falling from a height”, “hit by train”, or “pesticide poisoning” that were unnatural causes of death. Possible cases were suspected due to uncommon causes or premature death, for example, a 56–year–old male doctor who died because of “old age”, or a 42–year–old physician who died of a “skull fracture with internal bleeding”. The causes of suicidal deaths were categorized by ICD–9–CM codes (E950–E958). If the cause of death was not disclosed, code E958 was assigned.

The risk factors for suicide including demographic variables, medical specialties, practice settings and locality, were analyzed by univariate and multivariate logistic regression analysis using PASW Statistics software version 18 (SPSS Inc.). The expected number of cases of physician suicide was calculated as the sum of the annual product of physician all–cause mortality in each year from 2000 to 2013, multiplied by the age–, sex–, and year–specific suicide mortality ratio of the general population. Because the suicide mortality ratios of the general public in 2012–2013 are not yet available, the expected values of these two years were inferred from the data for 2011. PMR and 95% confidence intervals (CIs) for the physicians in Taiwan were calculated [27]. In addition, causes of suicide were examined in both genders and all age groups.


We excluded 5 records due to missing data. Among the remaining 1775 records of physician deaths between 2000 and 2013, 1738 (97.9%) cases were male, 101 (5.7%) were aged 25–44 years, 350 (19.7%) 45–64 years, and 1324 (74.6%) more than 65 years (Table 1). There were no records of death in a physician under 25 years of age. Most of the deceased physicians were born in Taiwan (922, 51.9%) and China (849, 47.8%); the rest were born in Canada, Malaysia, and the Philippines. Over half of the cases (944, 53.2%) had received domestic medical school education; the rest were educated in foreign medical schools (174, 9.8%) or qualified though special exam (657, 37.0%). Three–quarters of the registered specialties were included in general⁄internal medicine. Aside from the retired subjects (799, 45.0%), 627 (35.3%), 273 (15.4%), and 76 (4.3%) physicians practiced in private clinics, district⁄regional hospitals, and medical centers, respectively. The majority of the physicians worked in northern and southern Taiwan, where the largest urban areas are located.