Looking at the evolution in rate and cause of death in a hospice, over a twenty-year period, gives important insight into the development of the HIV epidemic. In 1988, HIV infected persons admitted to this facility were suffering from a ravaged immune system that resulted in deaths caused by a spectrum of diagnoses defined as AIDS-related conditions. All admissions were for palliative care. At the time, AIDS was felt to be a hopeless diagnosis. Casey House embraced the caring and compassionate hospice care model—often demedicalizing the treatment of the patients, and as a result specific investigations were not made into the particular cause of death and many certificates simply state this patient died of “AIDS.” In contrast, by 2008 less than 20% of patients admitted to Casey House died during their stay. And thus, 80% were discharged, often following an in-patient stay which included investigations and treatment for HIV as well as management of medical comorbidities and mental health issues. This highlights a significant change in the type of care needed by patients and also the lived experience of HIV-positive individuals.
Over the 20-year period examined in this study, there were significant changes in both the primary cause of death and the characteristics of those who died. In 1988, all deaths were attributed to an AIDS-defining condition, while less than half were in the 2006–2008 time period. Malignancies were a common cause of death in both time periods, although the types of malignancy changed considerably. Although the population remained largely male and mostly among men who have sex with men, in the post-HAART period 10% of deaths were in women. Individuals who died in the post-HAART period also reported greater IV drug and recreational drug use and more individuals were identified as homeless and with a mental illness.
Surveillance data and other research reports comparing the pre-HAART and post-HAART era show similar changes in the demographics of the HIV epidemic and causes of death. Although HAART has significantly decreased mortality in HIV-positive individuals, mortality is still significantly higher than in the general population and there are subgroups that are at substantially greater risk such as injection drug users [5
]. Studies linking HIV/AIDS databases and cancer registries over approximately the same time period have shown a dramatic decrease in AIDS-defining malignancies and an increase in non-AIDS-defining malignancies [12
]. As experienced at Casey House, the most common cancers of the pre-HAART era were Kaposi sarcoma and non-Hodgkin lymphoma. Both are typically linked with low CD4 count and coinfection with viral agents. Although such AIDS-defining malignancies remain prevalent in the HIV/AIDS population in the post-HAART era, numbers of non-AIDS-defining malignancies have increased as much as 20% in the US [14
]. Between 2001 and 2005, the most common types of non-AIDS-defining cancers were lung cancer, anal cancer, liver cancer, and Hodgkin lymphoma [13
]. This is reflected at Casey House. The Swiss cohort study notes that certain non-AIDS-defining malignancies associated with smoking such as lip, mouth, pharynx, and lung cancers increased in the post-HAART era. In addition, cancers of the liver associated with Hepatitis B and/or Hepatitis C coinfection, as well as human papilloma virus-related cancers, such as anal cancer, were also increased. It was speculated that living longer with disease, combined with partial immune reconstitution, allowed long-latency cancers to progress and manifest [12
The studies make the point that people living with HIV/AIDS are at greater risk of cancer than the general population, and there is a great deal of speculation as to why this occurs. Increased cancer risk is believed to be a complex interaction, with multiple factors, including impaired immunity, co-infection with biologic agents carcinogenic to humans (such as HPV, EBV, and hepatitis), aging, HAART, and traditional risk factors for cancer, such as smoking and sun exposure [15
]. There is some evidence that prolonged immunosuppression—as assessed by CD4 nadir—is independently linked to increased incidence of non-AIDS-defining malignancies [16
]. In addition, the presence of HIV is linked to cytokine dysregulation and this may also play a role in increasing cancer risks [15
Certainly there are complex factors at play in the patients seen at Casey House, who may have as many as 6 medical comorbidities, including coinfection with hepatitis, as well as risk factors such as unstable housing, injection drug-use, and smoking (approximately 50% of our patients smoke in comparison to 21% in individuals over 12 in Canada [17
]). Individuals are now living for more than a decade with HIV and although some have maintained CD4 counts above 200, with suppressed virus, many individuals are not on HAART therapy or do not adhere to their medication regimes. It is interesting to note that patients with reconstituted immune systems remain at risk from cancer [18
]. And, as reflected in the broader epidemic, our patients include more marginalized individuals who have significant social determinants of health risk factors such as homelessness and mental illness and this association also places them at risk [6
]. In response, we need to adopt effective health promotion models for HIV-infected individuals and the health care response must be comprehensive and collaborative. Research and service initiatives should expand their focus to address medical preventive care including cancer screening and treatment initiatives such as smoking cessation, as well as social and psychosocial needs. These services and supports must be available in various settings to address the spectrum of needs, from case management and chronic symptom management to acute medical interventions and end-of-life care.
We acknowledge that this retrospective study has several important limitations. Data extracted from charts are susceptible to variations in reporting and diagnostic criteria and missing data are common. The changes in reporting practices are amplified in this study where data were collected from time-points 20 years apart in a disease that has been described for only 30 years. In 1988, many primary causes of death were written as “AIDS.” It is most likely that these individuals and deaths would have been evaluated and recorded differently today. This may underestimate the number of individuals who died of AIDS-defining malignancies and possibly other non-AIDS-related causes of death. In this study, death certificates and primary cause of death were identified for all individuals but data were incomplete for many clinical and demographic variables. Our ability to clinically describe these patients is also limited by the state-of-knowledge and treatments available in 1988. Many tests routinely issued to HIV-positive individuals today, such as CD4 counts and viral loads, either did not exist or were not widely used in 1988. Hepatitis C was not yet identified [20
]. These findings are strengthened however by the complete inclusion of deaths that occurred at a single institution during the two periods of study. The description of all patients admitted in 2008 provides further context to the changes in the epidemic and the health care services provided. However, because Casey House is not an acute care hospital our study will not accurately represent the occurrence of certain types of death such as cardiac arrest, suicide, and trauma. It is also possible that individuals were differentially referred to Casey House in the two time periods as clinical knowledge and stigma of the disease evolved.