Concerns over the accuracy of South Africa's official mortality statistics have been amplified by a recent South African Medical Research Council (SAMRC) study. The study highlights the need to revisit HIV deaths in SA. This is given the potential discrepancies in reported data.
The research suggests a dramatic undercounting of HIV-related deaths. These may be potentially four times higher than figures reported by Statistics South Africa (Stats SA). This raises urgent questions about public health data and strategies regarding HIV deaths in SA.
A stark contrast
While health experts have long suspected that death certificates often mask the actual toll of HIV, the scale of the discrepancy revealed by the SAMRC study is surprising. This is especially true in the context of HIV deaths in SA.
Examining deaths from 2017/2018, the study concluded that HIV was the underlying cause of death in 23% of cases in a nationally representative sample. This figure starkly contrasts with Stats SA's data for the same period. They attributed only 5.7% of deaths to HIV.
Stats SA primarily relies on the cause of death listed on official death certificates. It's widely understood that HIV is frequently omitted due to stigma or other factors, even when it's the underlying cause. The SAMRC study, however, employed a more comprehensive approach to uncover the truth about HIV deaths in SA.
How the study found the hidden numbers
The SAMRC research team didn't just look at death certificates to arrive at its findings. They conducted a multi-phase investigation across 27 health sub-districts, analysing:
- Autopsy reports (forensic pathology records)
- Medical records from hospitals
- Interviews with next of kin (verbal autopsies using WHO questionnaires)
- Death certificates
Data was collected for over 26,000 deaths, with linked data allowing comparison between the study's assessed cause and the official death certificate cause. This comparison was possible for over 15,000 cases. A rigorous review process involving 49 medical doctors ensured accuracy.
The researchers found "poor agreement" between their findings and the official Stats SA data. The cause of death determined by the study matched the death certificate in only 37% of cases. Beyond HIV, the study also uncovered "severe under-reporting" of suicides.
Furthermore, the research underscored the devastating link between HIV and Tuberculosis (TB). It found TB responsible for 46% of deaths among people with HIV. Conversely, 63% of TB deaths occurred in individuals who also had HIV. Together, HIV and TB accounted for nearly 30% of all deaths in the study sample.
Challenging previous estimates
The SAMRC's 23% figure is significantly higher than even sophisticated estimates like those from the Thembisa model. Thembisa is South Africa's leading HIV model used by UNAIDS. It estimated that around 12% of deaths in 2018 were due to HIV. Study co-author Dr. Pam Groenewald acknowledged discussions with Thembisa researchers couldn't fully resolve the differences.
Potential factors contribute to the study's higher figures. They include under-representation of rural and child deaths, known to be under-registered generally. There is also the exclusion of deaths in private hospitals, where HIV-linked mortality is likely lower.
However, the study's strength lies in its clinical review of medical records, which includes CD4 counts and viral loads. This provides strong evidence for HIV's role in many deaths examined.
These findings also subtly challenge the narrative around mortality declines driven solely by antiretroviral therapy. They suggest the impact, while real, might be less pronounced than previously thought if the baseline HIV mortality was higher.
Public health implications
The study's authors stress the immediate need to reassess and strengthen HIV and TB strategies. Professor Debbie Bradshaw, a co-author and Chief Specialist Scientist at the SAMRC, said the study recommends strengthening case finding, follow-up, prevention, and treatment for HIV, AIDS and TB. The goal is to reduce mortality rates.
The research exposes fundamental weaknesses in South Africa's death certification system. Noting issues like incomplete documentation, the authors wrote that the findings highlight the need for improved record quality. There is also a need for adherence to testing guidelines within the medical community.
Improving death certification
Dr. Groenewald said that doctors often use vague terms like "retroviral disease" instead of explicitly stating HIV. This practice hinders accurate coding. She called for better training in death certification at medical schools and emphasized the legal and ethical obligation for doctors to record the actual underlying cause of death accurately.
Misconceptions about patient confidentiality regarding deceased individuals persist, but legal statutes require accurate reporting for public health purposes. Considering the discrepancies in HIV deaths in SA, improving transparency is vital.
Until these systemic issues are addressed, the study demonstrates that official mortality data from vital registration systems like Stats SA must be interpreted cautiously. This data should be cross-verified with other sources to guide effective public health interventions.