Table 5. Results of multilevel modelling of subject level variables on treatment for colorectal cancer within 90 days (yes/no) with levels subject, practice and health board.
| Odds ratio (95% confidence intervals)a N=1223 | |
|---|---|
| Individual variables | |
| Age (+1 year) | 1.3 (1.12–1.51) |
| Age squared (+1) | 0.998 (0.996–0.999) |
| Gender (male vs female) | 1.36 (0.98–1.89) |
| Palpable rectal mass (yes vs no) | 8.94 (2.48–32.13) |
| Palpable abdominal mass (yes vs no) | 4.32 (1.95–9.59) |
| Tenesmus (yes vs no) | 2.69 (1.02–7.09) |
| Abdominal pain (yes vs no) | 1.51 (1.06–2.14) |
| Frequency of consultations (+1 categoryb) | 3.25 (2.75–3.85) |
| History of anxiety/depression (yes vs no) | 0.47 (0.28–0.77) |
| On iron at presentation (yes vs no) | 3.77 (1.40–10.12) |
| Tumour site | |
| Right side colon to hepatic flexure | 1 |
| Transverse colon to splenic flexure | 2.27 (1.25–4.12) |
| Descending and sigmoid colon | 1.01 (0.63–1.60) |
| Rectum and recto-sigmoid | 1.40 (0.90–2.18) |
| Unspecified or overlapping | 1.26 (0.58–2.74) |
| Practice level variables | |
| Number of female GPs (+1) | 1.19 (0.99–1.44) |
| Variation (standard error) at | |
| Health board level 0.045 (0.055) | |
| Practice level 0.048 (0.112) | |
| No significant extra-binomial variation | |
| No significant effect of hospital type at health board level |
a
Odds ratio higher than unity means higher odds of being treated within 90 days. For interpretation of odds ratios see table 3, footnote 1.
b
Frequency of consultation categories: <0.5 consultations per month (from presentation to referral), 0.5–2.5 consultations per month, 2.5–10 consultations per month, >10 consultations per month.