Background:
We aimed to investigate the effect of comorbidity on cervical cancer survival for Australian women, and whether it contributed to the excess mortality of Indigenous Australian women compared to non-Indigenous women.
Method:
Cervical cancer cases diagnosed 2003-2012 in six Australian states/territories (covering 96% of the Australian female population) were linked to hospital inpatient records for the identification of comorbidity data. The weighted Charlson comorbidity index was calculated and categorised as a score 0 (no comorbidity), 1 or 2+. Kaplan Meier 5-year cause-specific survival probabilities were calculated by Indigenous status and comorbidity level. Flexible parametric survival models were used to model excess mortality by comorbidity score, for Indigenous and non-Indigenous women. The effect of comorbidity on cancer stage at diagnosis was also investigated for a sub-group of women from NSW, for whom staging data was available.
Results:
4,467 Australian women (4.4% Indigenous) with cervical cancer were included. Indigenous women were younger and had more comorbidity at diagnosis (score ≥1 24% vs. 10%, respectively). Indigenous women had lower survival than non-Indigenous women (60% vs. 77%). After adjustment for study factors, increased comorbidity was clearly associated with increased mortality for non-Indigenous women, but not for Indigenous women. Stratified analysis showed that the survival disparity between Indigenous women and non-Indigenous women was contained to those without comorbidity (HR 2.5, 95%CI 1.9-3.4). Indigenous and non-Indigenous women with comorbidity had similar survival.
Conclusion:
The difference in survival for Indigenous and non-Indigenous women without comorbidity warrants further investigation and may reflect differences in how health services engage these women.