Table 3. Results of multilevel modelling on treatment within 30 days (yes/no) with variation at patient, hospital of referral and health board level for people with breast cancer.
|
Odds ratio (95% confidence intervals)a |
||
|---|---|---|
| All referralsb N=1097 | Referred by GP N=822 | |
| Individual variables | ||
| Age at incidence (+10 years) | 1.16 (1.05–1.28) | 1.12 (0.99–1.26) |
| No. of symptoms | ||
| 0 (i.e. screen detected) | 11.92 (5.31–26.72) | |
| 1 | 1 | 1 |
| 2 | 1.43 (0.95–2.16) | 1.62 (1.05–2.51) |
| 3+ | 1.97 (1.07–3.64) | 2.39 (1.26–4.52) |
| Breast lump (yes vs no) | 3.16 (1.48–6.71) | 5.63 (2.20–14.42) |
| Change in skin contour (yes vs no) | 1.71 (1.02–2.85) | 1.75 (1.03–2.97) |
| Lymphadenopathy (yes vs no) | 2.05 (1.06–3.99) | 2.26 (1.13–4.53) |
| Hospital variables | ||
| Hospital type | ||
| Cancer centre | 1 | 1 |
| General | 4.18 (2.15–8.13) | 5.09 (2.64–9.79) |
| Community | 2.86 (1.45–5.66) | 3.63 (1.81–7.28) |
| Private | 13.38 (4.34–41.22) | 15.63 (5.09–47.95) |
| Variation (standard error) at: | ||
| Practice level 0.000 (0.000) | Practice level 0.000 (0.000) | |
| Hospital level 0.129 (0.096) | Hospital level 0.104 (0.094) | |
| Health board level effectively zero | Health board level effectively zero | |
a
Odds ratios higher than unity mean higher odds of being treated within 30 days. The values of odds ratios (and their differences from unity) provide an indication of effect sizes. The magnitude of odds ratios should be interpreted along with the unadjusted means in the previous tables.
b
Includes GP referrals, patients with cancers detected by breast screening and others.