Providing quality care for the body and mind


  • P. S. F. Yip,

    Corresponding author
    1. HKJC Centre for Suicide Research and Prevention, The University of Hong Kong, Hong Kong SAR, China
    • Correspondence: Paul Yip, HKJC Centre for Suicide Research and Prevention and Department of Social Work and Social Administration, The University of Hong Kong, 2/F, The Hong Kong Jockey Club Building for Interdisciplinary Research, 5 Sassoon Road, Pokfulam, Hong Kong SAR, China.

      (fax: (852)2549 7161; e-mail:

    Search for more papers by this author
  • S.-S. Chang

    1. HKJC Centre for Suicide Research and Prevention, The University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author

It is a pleasure to comment on Qin and colleagues’ excellent work on suicide risk amongst patients hospitalized due to physical illness. The authors have made good use of the unique and detailed registration systems in Denmark to be able to investigate suicide risk in relation to physical illness across a broad range of illnesses. This kind of study may not be possible in most of countries, including many developed countries, which do not have such comprehensive databases for the whole population. The findings provide empirical evidence of the relevancy of suicide prevention amongst the physically ill. Furthermore, the authors have quantified the potential effect of suicide prevention by looking at the population attributable risk (PAR), which is an estimate of the suicide reduction if the risk can be removed. A 24% PAR estimate suggests that clinicians indeed could be an effective gatekeeper for preventing suicide in the patient population, although it is difficult if not impossible to remove or reduce the suicide risk amongst all patients with physical illness.

The analysis of the study can be further enhanced if the age effect should be explicitly estimated. Suicide rate in men is about 2–4 times that in women [1]. Usually, older males had a relatively higher rate [2]. The presence of physical illness is also age and sex dependent. Chronic but not necessarily terminal illnesses are frequently found in people who die from suicide [3, 4] and these illnesses tend to be more common in old adults. This suggests that older men with physical illness would be a good target to prevent suicide. However, interestingly, the study shows a stronger effect of physical illness on suicide in females than males, as well as a larger PAR in women than men (32% versus 21%). This is in parallel to findings from a recent study based on data from UK primary care practices [5]. The UK study also indicates that increased risk of suicide is confined to young female patients (<50 years) and older women with multiple physical diseases. Further studies are needed to identify sex/age groups of physically ill patients who demonstrate the greatest risk of suicide and potential for prevention.

Furthermore, we feel that the role of mental illness on the association of physical illness with suicide deserves some further discussions. First, the study shows that the association weakens to some extent after controlling for history of contact with psychiatric services, including both outpatient and inpatient cares (unadjusted IRR = 2.13 versus adjusted IRR = 1.65). This attenuation is most prominent amongst the groups with likely the most severe illnesses, indicated by multiple hospitalizations and multi-morbidity. This may be partly due to the bias where more hospitalizations or more medical conditions give physicians more opportunities to detect psychological symptoms and refer the patients to psychiatric services. Second, perhaps more importantly, mental disorders may be a moderating factor on the pathway from physical illness to suicide or a confounding factor, which is associated with both physical illness and suicide. Pain, disability and physical disintegration caused by physical illnesses, as well as the traumatic experiences associated with the diagnosis of serious illnesses with poor outcomes such as cancer [6], can lead to emotional distress and depression and thus increased risk of suicide [4]. The UK study based on primary care data shows increased risk of suicide amongst physically ill patients but the association can be mostly explained by clinical depression [5]. One possible way to further investigate the role of mental illness on the link between physical illness and suicide is to examine the temporal relationship between hospitalizations due to physical illness and the contact with psychiatric services. Third, although the study shows evidence for an association of physical illness with suicide even after adjusting for psychiatric history, there is still possible residual confounding because many people with mental illness did not seek medical help. If this is substantial, the PAR would be an overestimate. Finally, the study shows that hospitalizations due to ‘symptoms, signs and abnormal conditions’ are associated with two times increased risk of suicide and contribute to a substantial proportion of PAR as they are relatively common amongst suicide cases (19%). These conditions may indicate undiagnosed psychological symptoms or disorders, as suggested by the authors, and the inclusion of them would lead to an overestimated PAR for physical illness.

In view of the exceptional high risk for inpatients (IRR = 13.4) and patients discharged within 7 days (IRR = 10.7), it might be tempting to develop a suicide risk or depression screening tool for these groups of patients based on these results [4]. To address excess risk of suicide amongst these patients, prevention strategies can also be developed to improve environmental safety within the hospital compound and the continuity of care after discharge in the first week. Prevention effect could also focus on patients with multiple hospitalizations and multiple comorbid physical illnesses, which were shown to be associated with a greater increase in suicide risk than patients with less episodes of hospitalizations and comorbid conditions in the study. Education and training in the detection and management of depression and suicide risk for a wide range of specialist clinicians treating patients with different physical illnesses may be beneficial, as the study showed increased risk of suicide across different categories of diseases examined. Clinicians should adequately treat pain and depression as well as pay attention to the safety of prescribed medications, such as analgesics and antidepressants, as they may be used in suicidal attempts. A psychiatric liaison team may also effectively identify patients with suicidal risk. More studies are needed to investigate risk and protective factors associated with suicides by patients with physical illness and a better understanding of these factors will inform prevention strategies. These factors may be context-specific as the patterns of patient care practices vary across different health care systems in different societies. Studies in different health care systems are therefore needed to inform context-specific prevention strategies.

Hence this is a wake-up call for clinicians that a holistic and multi-sectoral approach to patient care should be adopted. The compartmentalization and disconnection in medical care services is a barrier to the promotion of holistic care. Treating physical illness is important as it should be; however, neglecting psychological well-being is a serious concern. Appropriate psychiatric liaison care is also of great importance. To improve safety in hospital environments would involve collaborations with the hospital management team. For the continuity of care after discharge, it requires much more collaborative efforts working with relevant stakeholders in patients’ recovery process, for example, family members, liaison hospital staff and follow-up services. It is ironic if expensive cares are given to patients to treat their physical illnesses, and then they choose to take their own lives because of the distress, pain or fear associated with physical illness or the treatment. Suicide prevention is everyone's business and certainly clinicians would be effective gatekeepers.

Conflict of interest statement

No conflict of interest was declared.