Population-based utility scores for HPV infection and cervical squamous cell carcinoma among Australian Indigenous women

Abstract

Objective Working in partnership with Indigenous communities in South Australia, we aimed to develop, pilot test and estimate utility scores for health states relating to cervical cancer screening, precancer, and invasive cervical cancer and precancer/cancer treatment among Indigenous women. Methods Development and pilot testing of hypothetical cervical cancer health states, specifically through the lens of being an Indigenous Australian woman, was done with an Indigenous Reference Group in conjunction with five female Indigenous community members. Six health states were developed. These included: (1) Screened: cytology normal; (2) human papillomaviruses (HPV) positive with cytology normal; (3) low grade cytology (LSIL);(4) high grade cytology (HSIL); (5) early stage cervical cancer and; (6) later stage cervical cancer. Utility scores were calculated using a two-stage standard gamble approach among a large cohort of Indigenous Australian women taking part in a broader study involving oral HPV infection. The mean and standard deviation (SD) of the rank, percentage of respondents with a utility = 1 (perfect health) and utility score of each health state was summarised. Mean (SD) and medians and inter-quartile range (IQR) over 12 months and lifetime duration were calculated. Potential differences by age and residential location were assessed using the Wilcox Sum Rank test. Results Data was obtained from 513 Indigenous women aged 19+ years. Mean utility scores were higher for the four non-cancer health states than for invasive cervical cancer states (p-values <0.05). Lower mean utility scores were observed for late stage cervical cancer, with 0.69 at 12 months and 0.70 for lifetime duration (Intra-class correlation coefficients = 0.425). Higher utility scores were observed for the four non-cancer health states among non-metropolitan participants (ranged from 0.93 to 0.98) compared with metropolitan participants (ranged from 0.86 to 0.93) (p-values<0.05). Conclusion Among a large cohort of Indigenous Australian women, the reduction in quality of life (which utilities reflect) was perceived to be greater with increasing severity of cervical cancer health states. There were differences observed by geographic location, with positive cervical screening and precursor cancer-related quality of life being much higher among non-metropolitan-dwelling participants. These utility values, from one of the largest such studies ever performed in any population will be uniquely able to inform modelled evaluations of the benefits and costs of cervical cancer prevention interventions in Indigenous women.