In this nationwide retrospective cohort study matched by propensity score, stroke patients with FPC had the following characteristics compared to stroke patients with MPC: lower 30-day mortality; lower risk of pneumonia, septicemia, and acute renal failure; and shorter length of intensive care. We also found that FPC was associated with a reduced number of adverse events after stroke admission across a variety of conditions (e.g., patient age, medical utilization condition, physician age, and hospital volume) and patients’ severity of illness (e.g., CCI score and medical condition).
According to previously published studies, several factors influence the outcomes of stroke patients. The type of stroke is an important factor associated with stroke outcomes [19]. Patients’ baseline characteristics, including age, gender, age, sociodemographic data (level of urbanization and low-income status), and medical conditions (such as hypertension, diabetes, mental disorder, ischemic heart disease, COPD, cancer, heart failure, hyperlipidemia, liver cirrhosis, renal dialysis, and Charlson comorbidity index), were considered traditional factors that impacted stroke outcomes [20,21]. Hospital characteristics (hospital volume) and medical resource utilization (number of hospitalizations and emergency visits) might also influence the outcomes of stroke patients [7,22]. The characteristics of medical institutes might also play a role in the impact on the outcomes of stroke patients [8,23]. Physician age and division were also analyzed in our study. All the factors listed above could potentially influence the results; thus, we used the propensity score-matched pair model to reduce such cofounding effects in this study. We further performed multiple logistic regressions to adjust for residual confounding bias.
In addition to the factors listed above, physician gender may play an important role in the influence of stroke patients’ outcomes. Here, we propose a possible hypothesis to explain the findings of this study. First, in traditional Chinese culture, women are considered to be the main family caregiver, while men are expected to work hard and earn money to support their families. Therefore, male physicians had better success in advancing in their careers than did female physicians [24]. In work settings, male physicians spend much more time playing several roles, including those focused on teaching, research, administration management, and social activities. It is possible that male physicians might have less time than female physicians to dedicate to patient care. In the work setting of Chinese hospitals, female physicians are minorities, and their male colleagues sometimes help to do patient care. It is reasonable to hypothesize that female physicians might have fewer patients with relatively uncomplicated cases.Second, physician gender might affect his or her workload, and thus, the workload burden might play a role in stroke patients’ outcomes. In the United States, female physicians might have a lighter workload so that they can have more time with individual patients [25]. In Taiwan or Asian countries, limited information is available regarding this indicator. Given the previous description of traditional Chinese culture, it is reasonable to assume that Taiwanese female physicians might have a lighter workload and more time to care for individual patients.Third, physician gender might influence their provision of medical care to their patients. Female physicians were shown to have better communication than male physicians. Female physicians’ communication could be considered more patient-centered, focused on building an active partnership, having positive and emotionally intelligent conversation, providing psychosocial counseling and opportunities to ask questions, and allowing for a longer length of time for visits and consultations [26,27]. Female physicians also showed more empathy to their patients [28].Fourth, female physicians were more proactive in screening and prevention [29-31]. Patients of female physicians were significantly more likely to receive care that was consistent with guidelines and were more likely to seek help from other specialists [32,33].Finally, female and male physicians might have distinct methods of evaluating risk and making decisions [34]. In earlier reports, breast cancer patients might have different suggestions and medical advice for surgical and adjuvant radiation therapy [35]. All the previously mentioned factors might contribute to differences in the outcomes of stroke patients based on physician gender.Our research showed that stroke patients with FPC had better outcomes during the index hospitalization. Inconsistent results were observed in other fields of medical specialties in previous studies [9,11,12]. In obstetric nulliparous women who underwent a trial of labor, the outcome was similar among physicians of both genders [11]. In patients who received nonelective surgeries, including a wide range of surgeries (such as orthopedic, urologic, and general surgery), postoperative mortality was not associated with the surgeon’s gender [12]. In contrast to surgical patients, elderly inpatients receiving medical care from female internists had better outcomes (including mortality and readmission rates) than those who received care from male internists [9]. Patients with chronic conditions, such as diabetes, had a comprehensive preferred quality of care when they received their care from female physicians. Patients of female physicians had significantly superior progress in approaching the optimal levels of HbA1c, LDL cholesterol, and blood pressure [10]. Further studies are necessary to determine why such gender disparities emerged.
Some study limitations should be noted when the results of this study are interpreted. First, our insurance-based study lacks information on laboratory examinations, image findings, patient lifestyle (such as alcohol consumption, smoking habits, body mass index and physical activity levels) and stroke severity (such as measurements from the National Institutes of Health Stroke Scale or the Barthel Index). Second, the residual confounding bias could not be excluded from this study, although we used matching methods by propensity score to balance the baseline characteristics between stroke patients receiving care from male and female physicians. Third, our results were limited to 30-day in-hospital mortality and associated adverse events. Whether such a difference could extend to a longer period was not determined and requires further investigation. Additionally, patients’ neurological and functional recovery was not evaluated in this study.