Some text in this chapter has been reproduced with permission from Hunter et al.43 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.
Demographic characteristics for the prison and general population cohorts are reported in Table 4. Patients diagnosed with cancer in prison were less likely to attend outpatient appointments (OR 0.31 95% CI 0.25 to 0.39) and have planned inpatient attendances (OR 0.75 95% CI 0.64 to 0.88) than the matched cohort of patients diagnosed in the general population (Table 5). Among all patients who had any outpatient appointment, 20.8% (n = 275) of people diagnosed in prison had a ‘did not attend’ recorded for an outpatient appointment compared with 13.0% (n = 448) of the matched general population cohort. This represents a significantly higher likelihood of a ‘did not attend’ for patients diagnosed in prison (OR 1.79 95% CI 1.51 to 2.11).
| Place of diagnosis | p-valuea | ||||
|---|---|---|---|---|---|
| General population | Prison | ||||
| n | % | n | % | ||
| Patients | 4326 | 100 | 879 | 100 | |
| Sex | 0.94 | ||||
| Men | 4082 | 94 | 830 | 94 | |
| Women | 244 | 6 | 49 | 6 | |
| Age category (years) | 1 | ||||
| 18–20 | 29 | 1 | 6 | 1 | |
| 21–24 | 57 | 1 | 12 | 1 | |
| 25–29 | 109 | 3 | 26 | 3 | |
| 30–39 | 341 | 8 | 70 | 8 | |
| 40–49 | 693 | 16 | 140 | 16 | |
| 50–59 | 1078 | 25 | 216 | 25 | |
| 60–69 | 1144 | 26 | 231 | 26 | |
| 70–79 | 723 | 17 | 146 | 17 | |
| ≥ 80 | 152 | 4 | 32 | 4 | |
| Stage | 1 | ||||
| I | 1014 | 23 | 205 | 23 | |
| II | 534 | 12 | 108 | 12 | |
| III | 676 | 16 | 138 | 16 | |
| IV | 1262 | 29 | 253 | 29 | |
| Missing | 840 | 19 | 175 | 20 | |
| Diagnosis year | 1 | ||||
| 2012 | 505 | 12 | 102 | 12 | |
| 2013 | 662 | 15 | 135 | 15 | |
| 2014 | 697 | 16 | 142 | 16 | |
| 2015 | 775 | 18 | 157 | 18 | |
| 2016 | 867 | 20 | 176 | 20 | |
| 2017 | 820 | 19 | 167 | 19 | |
| Ethnicity | < 0.001 | ||||
| White | 3737 | 86 | 698 | 79 | |
| Mixed | 19 | 0 | 14 | 2 | |
| Asian and Chinese | 134 | 3 | 13 | 1 | |
| Black | 101 | 2 | 40 | 5 | |
| Other | 63 | 1 | 27 | 3 | |
| Missing | 272 | 6 | 87 | 10 | |
| Charlson comorbidity score | 0.07 | ||||
| 0 | 3622 | 84 | 708 | 81 | |
| 1–2 | 518 | 12 | 124 | 14 | |
| ≥ 3 | 186 | 4 | 47 | 5 | |
| General population (n = 4326) | Prison (n = 879) | ORa (95% CI) | |
|---|---|---|---|
| Outpatient | |||
| Proportion (%) | 93.39 | 82.25 | 0.31 (0.25 to 0.39) |
| Attendances,b mean (SD) | 11.97 (11.90) | 8.86 (10.96) | |
| Elective (planned) inpatient care | |||
| Proportion (%) | 35.99 | 29.82 | 0.75 (0.64 to 0.88) |
| Attendances,b mean (SD) | 1.38 (0.93) | 1.41 (1.02) | |
| Bed-days,b mean (SD) | 8.25 (10.78) | 8.61 (11.08) | |
| Day cases | |||
| Proportion (%) | 45.28 | 36.86 | 0.69 (0.60 to 0.81) |
| Attendances,b mean (SD) | 4.51 (5.55) | 3.63 (4.15) | |
| Emergency inpatient care | |||
| Proportion (%) | 16.10 | 14.79 | 0.90 (0.73 to 1.11) |
| Attendances,b mean (SD) | 1.32 (0.68) | 1.32 (0.72) | |
| Bed days,b mean (SD) | 13.92 (19.03) | 17.22 (21.77) | |
| Any inpatient care | |||
| Proportion (%) | 69.05 | 60.18 | 0.68 (0.58 to 0.79) |
| Attendances,b mean (SD) | 3.99 (5.12) | 3.25 (3.76) | |
| Bed-days,b mean (SD) | 10.49 (15.19) | 10.73 (15.38) | |
Based on link tests, Akaike information criteria, log-link and negative binomial family were chosen for all general linear models to test for significant differences in costs. Six months from diagnosis and 31 days prior to the diagnosis costs for secondary care for patients diagnosed in prison are significantly lower than for the matched cohort diagnosed in the general population (−£1216, 95% CI −£1638 to −£795), with outpatient and planned inpatient care costing significantly less for patients diagnosed in prison (Table 6). Emergency care costs £497 (95% CI £375 to £619) more per patient diagnosed in prison. When the cost of bed-watches and escorts is added to the total cost of secondary care cancer treatment, patients diagnosed in prison cost significantly more per patient than the matched general population cohort.
| General population (n = 4326) | Prison (n = 879) | Differencea (95% CI) | |
|---|---|---|---|
| (1) Outpatient | |||
| Cost,b mean (SD) | 1651 (1693) | 1209 (1541) | |
| Adjusteda total cost, mean (SE) | 1544 (24) | 981 (33) | −563 (−643 to −483) |
| (2) Elective (planned) inpatient care | |||
| Cost,b mean (SD) | 9382 (9902) | 9424 (10,633) | |
| Adjusteda total cost, mean (SE) | 3414 (57) | 2919 (101) | −495 (−715 to −275) |
| (3) Day cases | |||
| Cost,b mean (SD) | 1516 (1639) | 1224 (1160) | |
| Adjusteda total cost, mean (SE) | 685 (11) | 455 (16) | −230 (−267 to −193) |
| (4) Emergency inpatient care | |||
| Cost,b mean (SD) | 7473 (12,840) | 8766 (12,729) | |
| Adjusteda total cost, mean (SE) | 1153 (21) | 1650 (63) | 497 (375 to 619) |
| Total inpatient care (2–4) | |||
| Cost,b mean (SD) | 7619 (11,144) | 7589 (10,803) | |
| Adjusteda total cost, mean (SE) | 5234 (82) | 4601 (157) | −634 (−977 to −291) |
| Total health care costs (1–4) | |||
| Adjusteda total cost, mean (SE) | 6784 (106) | 5568 (190) | −1216 (−1638 to −795) |
| Total healthcare costs (1–4) including escorts and bed-watches | |||
| Adjusteda total cost, mean (SE) | 6784 (106) | 17,085 (582) | 10,301 (9145 to 11,456) |
The results for the sensitivity analyses are reported in Appendix 1. If patients with unknown NHS number are excluded from the analysis, the results remain the same as for all other analyses.