For older male outpatients with schizophrenia, our analyses support a modest longitudinal association over an 8-year period between a greater scope of guideline-concordant care and survival. However, patients in the group who received the least comprehensive guideline-concordant care had better adjusted odds of survival than those receiving high or moderate levels of care, although the benefit relative to those with high-consistent care was small. This group was characterized by lower comorbidity, older age (but a larger age range), and fewer VA priority 1 veterans (those with a high level of disability), thus apparently less need for care. The mechanisms by which this survival benefit might accrue include access to care being selectively used, emphasis on prevention and early disease management being common in primary care, and a focus on necessary care [30
]. Possibly these patients also avoided excess medical care that might itself carry increased risk to their well-being through iatrogenic mechanisms. Moreover, patients in the poor-decreasing care group had less comorbidity, which is known to account for half the excess mortality of persons with schizophrenia [31
]. Their relatively lower rates of cardiovascular diagnoses may account for some of the findings and the lower levels of care.
Interestingly, this sample of veterans with schizophrenia received on average only 4.8 out of the 11 care types examined. The patient's burden of illness translates into the healthcare system's challenge of providing appropriate preventive services; difficult care coordination for patients with the diminished self-care capacity or cognitive deficits common with schizophrenia may account for this modest level of guideline-concordant care [32
]. Least common were visits for nicotine dependence, weight management, drug/alcohol dependency infectious diseases, eye care, and group psychotherapy. Over time, possibly due to aging, the percentage of patients receiving cardiovascular and eye care tended to increase, which is appropriate given the age-related nature of cardiovascular and eye disorders. The failure to address weight management is unfortunate, as this common condition potentiates much cardiovascular disease. Other research on VA patients with schizophrenia noted no disparities in receipt of obesity care practices, but did note a generally low level of weight management [33
Of note, the proportion of patients receiving mental health treatment (psychiatric consults for medication management, and individual, family, and group therapy) gradually decreased over time. Regarding medication management, this may indicate diminishing use of antipsychotic agents in later life. A 2004 review suggested that older persons with schizophrenia do not need fewer medications, but a recent analysis suggests that they may require lower doses or antipsychotics or none at all [34
]. Nevertheless, patients who used antipsychotics consistently had a longer survival than those with poor or inconsistent use, seemingly in line with the report by Maher and Theodore, who found that appropriate antipsychotic use was not associated with increased mortality [36
]. The results from liver and renal function tests, often used to monitor adverse drug effects, were generally high [37
]. Families involved in the care of these patients should be mindful of the need to work closely with care providers regarding medication options (for example, dose adjustment, possible medication changes to limit weight gain). Psychological counseling, particularly family counseling, can improve illness insight and thence treatment adherence, and also improve family functioning [38
]. The declines in mental health counseling merit investigation into the client and family need for and benefit from these services.
Patients who fail to come in for outpatient care for 12 months or more [17
] experience heightened risk of death, whether through increased risk due to heart disease [39
], adverse effects of antipsychotics on cardiovascular functioning and metabolic regulation [40
], or increased risk of liver or renal disease [16
]. Over time, the proportion of patients receiving renal/liver function care and psychiatric medication management in the moderate-consistent and poor-decreasing care groups gradually declined. Receiving liver function screening, especially for patients with liver disease, is associated with appropriate treatment and longer survival [42
]. Insufficient monitoring would provide fewer opportunities to make adjustments in medication regimens including hepatotoxic medications, and patients are likely to continue using psychotropic medications that exacerbate liver problems if poor liver functioning goes undetected. Accordingly, providers and family caretakers should be vigilant when patients decrease their healthcare visits over time or miss appointments.
As this is perhaps the first study to examine the association of comprehensive guideline-based care with survival for older male patients with schizophrenia, our findings suggest that greater adherence to screening and treatment recommendations is crucial to patient survival, and may reduce premature mortality. Additionally, early detection and treatment of incipient problems is likely to produce greater returns in terms of quality of life for older patients with schizophrenia, and to reduce the strain of caretaking for the older patient's social network. It is important for mental and primary healthcare practitioners to coordinate care efforts and improve overall adherence to these treatment guidelines. These patients may benefit from aggressive outreach and use of a patient-centered medical home to improve care coordination and treatment adherence.
Limitations and strengths
This study relied on administrative data, thus severity ratings of psychiatric or medical illnesses were not available and nor were quality-of-life assessments. Medicare and other out-of-system data were unavailable, thus any such services received were not captured by our study methods [23
]. Future research linking VA with Medicare data to assess guideline-concordant care could determine the extent of bias introduced in our results by reliance on a single healthcare system for those persons covered by Medicare. For example, patients in the poor-decreasing care cluster may have been more likely than others to seek healthcare services outside the VA after the baseline year, as many were reaching the age of 62 (early eligibility) or 65 (usual eligibility) years during the period of observation. On the other hand, patients with serious mental illness tend to stay within the VA [43
]. Although a high proportion of patients received family counseling, it is uncertain whether this translated into evidenced-based family services for schizophrenia, considering that previous studies have identified low rates of such care [44
]. Of note, the present study focused on male veterans age 50 years or older, therefore results cannot be generalized to women or younger patients. VA patients tend to have lower incomes and more health problems than the general population in the USA, which may affect their need for care and the breadth of healthcare options available to them [45
The strengths of the study include its large scale and longitudinal design, which enabled us to follow patients over an 8-year period, and an innovative approach mapping the treatment recommendations from APA guidelines onto the indicators of care received. The study addressed a crucial and serious healthcare problem, namely, premature mortality among patients with schizophrenia. We were careful to include preventive care, such as screenings for common medical and psychiatric comorbidities, as well as treatment for these conditions. A further strength was the use of cluster analysis, which enabled a practical and interpretable comparison of patient groups characterized by different patterns of care over time.
Future research on the effects of preventive care for individuals with schizophrenia is needed to replicate and extend the results of this study. Studies are needed in other healthcare systems, and including people who lack access to consistent sources of care. Also of value would be an examination of the relative importance of the types of recommended care in reducing premature mortality. For example, through family counseling, family members of patients with schizophrenia may be more likely to assist with medical adherence and coordination. This may be especially important when patients with schizophrenia require complex coordinated care, which may be difficult to manage.