Acute appendicitis is a common surgical emergency. A growing inventory of medical literature has used differential rates of ruptured appendix as an indicator of a potential barrier to the accessibility of medical care [1–7]. Although symptomatic acute appendicitis is not preventable, timely treatment of symptomatic acute appendicitis can help to prevent appendiceal perforation, which is associated with severe morbidity and elevated mortality[1, 8–10]. According to previous literature, the strongest predictor of perforation is postponed surgery in excess of 12 to 24 hours after the symptoms of appendicitis first occured. Previous reports observed higher rates of appendiceal perforation among people without insurance, minorities, young children, the elderly and men [1–3, 7, 8, 11, 12]. However, none of these studies had systematically examined whether there was a difference between patients, with or without a mental disorder. In addition to direct physiological mechanisms, hindered access to necessary medical care may lead to a greater number of adverse outcomes of co-morbidity among patients suffering from mental illness [13–20].

Patients with a major mental disorder, such as schizophrenia or affective psychosis, may be at a higher risk of encountering hindered access to care for their physical diseases, possibly due to their impaired capacity to recognize or articulate the emerging medical illness [15, 18, 19, 21, 22]. Therefore, not only does the access of mentally ill patients to mental health services deserves policy attention, but their access to care for their physical diseases should also not be overlooked. For example, the typical consequences of schizophrenia, such as cognitive impairment, social isolation, suspicion and possible pain insensitivity may cause mentally-ill patients to be less likely to seek medical help, promptly report physical symptoms or follow prescribed treatment [23–29]. Also, physicians' lack of experience with mentally ill patients may lead to misdiagnosis or delayed diagnoses in certain cases. Some psychotic symptom such as somatic delusions may lead physicians whom are less experienced in treating mentally ill patients to overlook some of the physical symptoms presented. Furthermore, many psychiatric institutions may not have sufficient diagnostic equipment or surgical facilities to provide timely care for the physical ailments of their patients. Factors related to patient, physician and institution, may impede mentally ill patients, particularly schizophrenic patients, from receiving timely care [14, 15, 20, 22, 30].

Although only anecdotal evidence and a few reports based on small samples suggest that schizophrenia patients may have a higher rate of perforated appendicitis [31], there is no large-scale study available in the medical literature on this topic. Moreover, in addition to providing insurance [32, 33], a large variety of extensive efforts are devoted to assure access to mental health services by mentally ill patients by addressing cultural, linguistic, and other barriers. However, we should not limit our attention to these patients' access to care for mental diseases, but also their access to care for physical diseases. The implementation of the National Health Insurance (NHI) in Taiwan in 1995 successfully reduced financial barriers to access care for all including the mentally ill. Nevertheless, the purpose of this study was to investigate whether disparities in access to care for appendicitis still exist between patients with and without mental illness, and more specifically, between patients with schizophrenia and other major mental illness. Our work contributes to the existing literature by conducting a first population-based study to systematically examine the association between schizophrenia, affective psychosis, other major mental illnesses, and appendiceal perforation. Although our findings may not lead to important hints about the etiology or pathophysiology of mental illnesses, they can help to identify disparities so that we can devise strategies to reduce these disparities in access to timely surgical care for physical illnesses between patients with and without mental illness, and between patients with different mental illnesses.

Table 1 describes the characteristics of the study sample according to the presence or absence of a ruptured appendix. A ruptured appendix occurred in 46.7 percent of the schizophrenic patients; in 43.4 percent of the patients with other major mental disorders; in 27.6 percent of the patients with affective psychoses; and in 25.1 percent of the patients with no major mental diseases. More ruptured cases were found among males and older patients. The rate of ruptured appendix decreased from 27.3% in 1997 to 22.9% in 2001. Overall, although the total number of symptomatic acute appendicitis increased from 18,778 to 20,625, the incidence rate remained relatively constant (approximately 0.1%) due to the natural population increase and the rising NHI enrollment rate during the study period.

After adjusting for age, gender, ethnicity, SES, and hospital characteristics, the odds ratio increased from 2.61 to 2.83 and schizophrenic patients were still the most likely among all of the patients with appendicitis to experience appendiceal rupture. The next most likely were patients with other major mental disorders, but the risk was no longer significant (OR: 1.58, 95% CI: 0.89–2.81) after taking patient characteristics, hospital characteristics, and years into consideration. One plausible explanation may be that as age is strongly associated with rupture according to previous research, age might be a most likely factor to reduce the odds ratio for the "other" group in the multivariate analysis. Also, after taking more variables into consideration, it might lose degrees of freedom in the multivariate model and hence the odds ratio might become statistically insignificant. Similar to the results of simple logistic analysis, after adjustment, patients with affective psychoses (OR: 1.15, 95% CI: 0.77–1.73) remained insignificantly associated with elevated ruptured appendix rates.

As for the other characteristics, the multivariate results were consistent with previous findings [1, 7–11, 40]. For adults, increasing age, being male, and lower SES were also significantly associated with a higher risk of perforation. In addition, after adjusting for other patient and hospital characteristics, aboriginals were at significantly higher risk of having appendiceal perforation than non-aboriginals.

This population-based study found that for all patients with mental illness, the strongest positive association between mental illness and ruptured appendix is for patients with schizophrenia, independent of other risk factors found in previous literature, such as age, sex, SES, ethnicity and hospital characteristics. Previous research indicates that, on average, schizophrenic patients have a 2 to 4 times higher mortality rate than the general population, although this excess mortality has been mainly associated with an elevated risk for suicides or traumatic injuries [23, 41–43]. One extreme example is from a paper by Munk-Jorgensen (2000), which indicated that although some other factors could also have contributed to the findings, the rate ratio for schizophrenic patients' admissions to somatic hospitals for acute and potentially lethal diseases is greater than 1, up to a maximum of 4.15 for severe heart failure [17]. Our finding of OR = 2.83 for ruptured appendix among schizophrenic individuals is consistent with previous reports of excess adverse outcomes among schizophrenic patients. Excess adverse outcomes from physical causes among patients with schizophrenia are a serious problem.

Furthermore, not all mentally ill patients are equally vulnerable. Compared to patients with other mental illness, a much higher rate of ruptured appendix was identified among schizophrenic patients in this study. This finding suggests a substantial gap in accessibility to necessary medical care, between patients with and without schizophrenia. The schizophrenic patient's possible insensitivity to pain, cognitive or social impairment, general or surgical practitioner's inexperience in treating schizophrenic patients, inadequate diagnostic or surgical facilities in psychiatric institutions, poor family support, other health system or psychosocial factors which may lead to misdiagnosis, delayed diagnosis and delay in appropriate treatments, could partially explain the excessively higher rates of ruptured appendix observed among schizophrenic patients compared to patients without a mental illness. In addition, some reported pain insensitivity in schizophrenic patients may make them particularly vulnerable to adverse consequences of physical diseases that are typically pain-causing [25–27]. Although some pain insensitivity in patients with schizophrenia has been reported in the literature and the phenomenon has been explained by several advanced theories [26, 28], some previous studies suggest that it is not well recognized or understood in general or surgical practice [26, 29, 44].

The large sample size and the robust adjustment for comprehensive patient and hospital characteristics have allowed this study to obtain a more precise estimate of the independent contribution of schizophrenia to the rates of perforated appendix. Controlling for hospital characteristics was done so as to reconcile a possible variation among physicians in diagnostic and reporting practices. Furthermore, we studied admissions for acute appendicitis only, rather than for appendectomy, in order to avoid including admissions for reasons other than acute appendicitis.

Several study limitations should be noted. First, misdiagnosis of acute appendicitis may lead to an underestimation of perforation rate. Misdiagnosis could be differential or non-differential between schizophrenic patients and others. According to the previous literature [45–47], the misdiagnosis rate of acute appendicitis ranges from 15% to 40%. The negative association between schizophrenia and acute appendicitis, and the under-reporting of medical illnesses in schizophrenic patients found in previous studies imply that overdiagnosis of acute appendicitis may be more prevalent in non-schizophrenic patients than in schizophrenic patients [23, 48–50]. Assuming an extreme differential case, where the reference patient group (who did not have any major diseases) had a 40% overdiagnosis rate and the schizophrenic patients had no misdiagnosis, our estimate of the odds ratio between schizophrenic patients and reference group (patients with no major diseases) would be an overestimation. On the other hand, if the misdiagnosis was non-differential, the odds ratio would be larger than the one observed in this study. Due to the lack of pathology, outpatient and emergency room data, we were unable to address the possible misdiagnosis issue in this study. Future research with more comprehensive data may help in this regard. Second, our reliance on the NHI major disease files for mental disorder case definitions may be a problem. The patients with a major disease/injury card of a listed major mental diseases, must have a confirmed diagnosis from a psychiatrist that long-term treatment is necessary. More than six months psychiatric treatment duration is required to validate the necessity of long-term treatment. Hence, because of these stringent criteria, the major disease file is likely to capture those patients with a chronic severe mental disorder. However, it may not have captured all patients with schizophrenia. Furthermore, the stigma associated with schizophrenia may discourage some schizophrenic patients from applying for a major disease card, so these patients may also not be captured by the major disease file. In addition, misclassification of patients with schizophrenia into other groups may also lead to an underestimation of the differences between patients with and without schizophrenia. Moreover, the sample size limitation restricts us to conducting more detailed interaction analyses such as the interaction between mental disorder and age. An insufficient sample size may result in the large standard error. Future research with a larger sample size may help to detect differential mental disorder impacts on ruptured appendicitis between different age groups. Finally, our proxies of SES were limited to individual-level insurable wages in the enrollment category and so they may not be comprehensive.

The findings of our study suggest that mentally ill patients are at a disadvantage for obtaining timely treatment for their physical diseases. Of patients with a major mental illness, schizophrenic patients may be the most vulnerable ones due to biological, health care system and psychosocial factors as discussed above. Atypical presentations in schizophrenic patients raise serious clinical concern for an increased risk of misdiagnosis regarding evaluation as well as delayed treatment. How to reduce disparities in physical health for the mentally ill patients is a complex issue and will need a multidimensional strategy to address the possible barriers among this vulnerable subpopulation. For example, a better integrated care delivery system such as case management programs for a closer integration of physical and mental health services of mentally ill patients would be of some benefit. It would allow physicians to have a better understanding of a patient's history and present status, which could help physicians distinguish between symptoms of serious illnesses and noise. In addition, a higher perforating rate among mentally-ill patients may also be related to patient factors and system factors. Therefore, educational interventions such as providing more health information or training to patients' families or caregivers may help to reduce delayed presentation for care due to the pain insensitivity or delayed personal recognition of the potentially-serious nature of symptoms. Organizational strategies may also help to facilitate or reduce possible delays in diagnosis or treatment processes. For mentally ill patients, not only are their mental health concerns important, their physical health issues are equally critical. Taiwan's experience suggests that improving health for mentally ill patients is more than a matter of finances. More attention is required on how some health care features are organized, as well as educational interventions. Both may help to reduce possible non-financial barriers and reduce the excess morbidity and complications for this vulnerable population, both in Asia and around the world.