Conclusion

The linear probability models created for the geographic variation analysis all accounted for differences in the distribution of various patient demographics and tumour characteristics between Cancer Alliances that might have otherwise confounded the main association under study. Despite such adjustment, geographic variation in treatment remained. Rather than representing real disparities in clinical practice, some of these geographic differences may be the result of residual confounding, such as differential routes to diagnosis between regions (unavailable at the time of analysis), variation in the proportion of patients who died before the primary course of treatment could be started or concluded, or geographic differences in patient frailty not captured through the Charlson comorbidity score.

The reported analyses did not consider privately funded treatments. Due to the absence of readily accessible private healthcare data, tumours in private patients may have been incorrectly assigned to the ‘no major surgical resection or chemotherapy’ category. Accordingly, the true proportion of tumours that received treatment will likely be higher than reported, with the additional possibility of differences in private treatment access between Cancer Alliances. If present, this will explain some of the variation observed between Cancer Alliances in analyses where tumours assigned to the ‘no major surgical resection or chemotherapy’ category are included. The authors are not aware of any data sources that allow for a reliable estimate of the degree to which this misclassification may have occurred.

Surgery was defined as the delivery of at least one major resection during the primary course of treatment. Major surgical resections do not encompass all surgical procedures delivered to ovarian tumours, excluding procedures such as diagnostic biopsies. If a broader definition of surgery were to be applied, treatment rates in these patient groups would be expected to exceed those described in this report. Major surgical resections were selected because they constitute the main surgical intervention for the treatment of ovarian cancer.

Charlson comorbidity scores were defined by linking each tumour to pre-defined comorbid medical conditions documented prior to the diagnosis of ovarian cancer. A list of the medical conditions considered and the associated score for each is described in Table 21 of the data downloads section. . Through a dependence on diagnosis coding within the inpatient setting, the Charlson comorbidity score may underestimate the burden of index-relevant comorbidity by missing diagnoses exclusively documented in outpatient or primary care settings. However, in the context of survival estimation, a comparison of Charlson comorbidity indexes derived for a fixed general population cohort of adults aged >20 years found that an index based on secondary care data performed at least as well as one that utilised primary care data. ¹

Beyond this, through its selection on a specific subset of chronic health conditions, the index may not reflect the full burden of all comorbid disease that may influence clinical decision making. For instance, Table 7 shows that 69.7% (n=898) of tumours in the cohort were assigned a Charlson comorbidity score of zero, representing tumours in patients without any record of another primary cancer in the cancer registry or a pre-defined comorbid medical condition documented within an inpatient setting prior to the diagnosis of ovarian cancer. That more than two-thirds of tumours in the cohort received a comorbidity score of zero sits at odds with the age profile of the ovarian cancer cohort, as well as research elsewhere for other tumour sites that has demonstrated a broad range of comorbid medical conditions. ² Nevertheless, the Charlson comorbidity index captures at least some of the variation in the probability of treatment, whereby tumours in patients with higher scores were reported as having lower probabilities of any treatment (Table 10), any surgery (Table 12) or any chemotherapy (Table 14).

Notes:

1. Crooks CJ, West J, Card TR. A comparison of the recording of comorbidity in primary and secondary care by using the Charlson Index to predict short-term and long-term survival in a routine linked data cohort. BMJ Open. 2015;5(6).
2. Fowler H and others. Comorbidity prevalence among cancer patients: a population-based cohort study of four cancers. BMC Cancer. 2020;20(1):1–15.

Finally, this report described geographic variations in treatment according to the Cancer Alliance of residence at the time of diagnosis. It is possible that some tumours may have received treatments from multiple Cancer Alliances over the course of treatment, which may have differed from the diagnosing Cancer Alliance. As the Cancer Alliance at treatment can vary over time and according to treatment type, Cancer Alliance at diagnosis was reported for simplicity. Comparison with data from 2016 to 2018 diagnoses published in the Treatment report is also complicated by the reconfiguration of Cancer Alliances.