In general, similar use of services was seen in the other health conditions as well. Increased use of a care worker day to day in both CFAS I and CFAS II was apparent in those with dementia and all other health conditions apart from Parkinson’s disease, where the number of people with dementia and Parkinson’s disease who used the service were too low to complete analysis (Additional file
1: Table S4–S11). Unpaid care was increased in CFAS II and not CFAS I for angina (Additional file
1: Table S4), Parkinson’s disease (Additional file
1: Table S5), breathing difficulties (Additional file
1: Table S8), transient ischaemic attack (Additional file
1: Table S9), heart attack (Additional file
1: Table S10), and hearing difficulties (Additional file
1: Table S11), whereas for hypertension (Additional file
1: Table S6) and anaemia (Additional file
1: Table S7) unpaid care was increased in CFAS I and CFAS II. In most other health conditions where there were the numbers to complete analysis, day-to-day paid help was decreased in both studies, rather than just in CFAS II (Additional file
1: Table S4-S11). Home help use in the 4 weeks prior to interview was consistently increased in both studies for the other health conditions (Additional file
1: Table S4–S11). Day centre use in the 4 weeks prior to interview was also generally increased in both studies, where it was possible to complete the analysis. Increased nursing service use in CFAS II but not CFAS I was also seen in anaemia (Additional file
1: Table S7) but, otherwise, use was similar between studies in other health conditions. Increased use of inpatient hospital services in both studies was only seen in breathing difficulties (Additional file
1: Table S8) and transient ischaemic attack (Additional file
1: Table S9), otherwise it was increased in CFAS I but not CFAS II for angina (Additional file
1: Table S4), hypertension (Additional file
1: Table S6), heart attack (Additional file
1: Table S10) and hearing difficulties (Additional file
1: Table S11).
The sensitivity analysis conducted showed that most estimates from the deprivation adjusted models were within CIs for the age- and sex-adjusted models for CFAS I and CFAS II (Additional file
1: Table S13–S24). The only exception was the estimate comparing the number of day-to-day services accessed by those with angina and dementia in comparison with angina alone in CFAS I (Additional file
1: Table S16). This does not change any of the conclusions, as the estimate became stronger, from IRR 1.6 (95% CI 1.1–2.3, age and sex adjusted) to IRR 2.9 (95% CI 1.8–4.8, deprivation adjusted).