Table 5.
Randomisation test across providers who saw differently gendered case presentations
|
Saw a woman presenting |
Saw a man presenting |
Adjusted odds ratio (regression estimate) | 95% CI | p value | |||
|---|---|---|---|---|---|---|---|
| Number of interactions | Correct management proportion | Number of interactions | Correct management proportion | ||||
| Case 1 | |||||||
| Case 2 (Women) | 50 | 0·70 | 161 | 0·63 | 1·08 | 0·55–2·11 | 0·82 |
| Case 2 (Men) | 44 | 0·66 | 165 | 0·66 | 0·91 | 0·43–1·94 | 0·81 |
| Case 3 (Women) | 39 | 0·33 | 83 | 0·25 | 0·83 | 0·27–2·55 | 0·74 |
| Case 3 (Men) | 48 | 0·25 | 220 | 0·31 | 1·22 | 0·55–2·72 | 0·63 |
| Case 4 (Women) | 37 | 0·22 | 95 | 0·13 | 0·65 | 0·14–2·91 | 0·57 |
| Case 4 (Men) | 46 | 0·17 | 310 | 0·10 | 0·45 | 0·10–1·94 | 0·28 |
| Case 2 | |||||||
| Case 1 (Women) | 54 | 0·39 | 45 | 0·47 | 2·46 | 0·86–7·05 | 0·093 |
| Case 1 (Men) | 189 | 0·49 | 201 | 0·45 | 1·07 | 0·56–2·04 | 0·84 |
| Case 3 (Women) | 9 | 0·33 | 1 | 1·00 | .. | .. | .. |
| Case 3 (Men) | 15 | 0·40 | 18 | 0·44 | 7·77 | 0·61–98·56 | 0·11 |
| Case 4 (Women) | 6 | 0·33 | 6 | 0·50 | 4·38 | 0·22–89·15 | 0·34 |
| Case 4 (Men) | 21 | 0·29 | 14 | 0·07 | 0·00 | 0·00–0·03 | 0·00048 |
| Case 3 | |||||||
| Case 1 (Women) | 41 | 0·39 | 52 | 0·42 | 2·13 | 0·54–8·38 | 0·28 |
| Case 1 (Men) | 105 | 0·45 | 253 | 0·43 | 1·28 | 0·61–2·66 | 0·52 |
| Case 2 (Women) | 9 | 0·44 | 13 | 0·69 | 10·72 | 1·03–111·07 | 0·047 |
| Case 2 (Men) | 1 | 1·00 | 17 | 0·76 | .. | .. | .. |
| Case 4 (Women) | 3 | 0·33 | 11 | 0·45 | .. | .. | .. |
| Case 4 (Men) | 13 | 0·23 | 29 | 0·31 | 1·58 | 0·46–5·37 | 0·46 |
| Case 4 | |||||||
| Case 1 (Women) | 37 | 0·43 | 43 | 0·40 | 0·62 | 0·15–2·60 | 0·52 |
| Case 1 (Men) | 101 | 0·48 | 336 | 0·35 | 0·66 | 0·31–1·39 | 0·28 |
| Case 2 (Women) | 6 | 0·83 | 18 | 0·72 | 2·64 | 0·14–48·29 | 0·51 |
| Case 2 (Men) | 6 | 1·00 | 14 | 0·50 | .. | .. | .. |
| Case 3 (Women) | 3 | 0·00 | 10 | 0·40 | .. | .. | .. |
| Case 3 (Men) | 12 | 0·58 | 28 | 0·39 | 0·49 | 0·13–1·90 | 0·30 |
For each case scenario, this table shows a test of balance across the providers who saw a man present that case and the providers who saw a woman present that case. For each other gender-case presentation, it assesses whether any significant difference exists between those two groups of providers. The table presents the N, mean correct management proportion, odds ratio, 95% CI, and p value for differences in correct management between those two groups. Reported odds ratios are logistic regression coefficients on the gender of the standardised patient, controlling for city, case scenario, and provider qualification, and standard errors are clustered at the health care facility level.