Table 6.
Numbers of patients with recordings and nurses’ responses to recordings that should have and did, or did not, trigger a response a in the first eight postoperative hours (per protocol analysis)
| Intervention arm | Control arm | ||||||
|---|---|---|---|---|---|---|---|
| Parameter | MEWS chart n = 36 patients | Existing chart n = 57 patients | P value | OR (df = 1) | 95% CI | ||
| Number (%) | Response triggered | Number (%) | Response triggered | ||||
| Respiratory rate recorded | 25 (69.4) | 2 (3.5) | <0.001 | 62.50 | 12.89-303.15 | ||
| Respiratory rate should have triggered a response | 20 (55.6) | 0 | 0 | 0 | 1.00 | ||
| Heart rate recorded | 36 (100) | 57 (100) | 1.00 | ||||
| Heart rate should have triggered a response | 12 (33.3) | 0 | 19 (33.3) | 0 | 1.00 | ||
| Oxygen saturation recorded | 6 (16.7) | 2 (3.5) | 0.03 | 5.5 | 1.05-28.95 | ||
| Oxygen saturation should have triggered a response | 1 (2.8) | 0 | 1 (1.8) | 0 | 1.00 | ||
| Systolic blood pressure recorded | 36 (100) | 57 (100) | 1.00 | ||||
| Systolic blood pressure should have triggered a response | 15 (41.7) | 3 (20.0) | 25 (43.9) | 1 (4.0) | 0.23 | 5 | 0.48-52.53 |
| Temperature recorded | 35 (97.2) | 54 (94.7) | 1.00 | 1.94 | 0.19-19.45 | ||
| Temperature should have triggered a response | 17 (47.2) | 0 | 22 (38.6) | 0 | 1.00 | ||
| Level of consciousness recorded | 33 (91.7) | 37 (64.9) | 0.004 | 5.95 | 1.62-21.84 | ||
| Level of consciousness should have triggered a response | 6 (16.7) | 0 | 7 (12.3) | 0 | 1.00 | ||
| Urine output recorded | 33 (91.7) | 51 (89.5) | 0.12 | 1.29 | 0.30-5.54 | ||
| Urine output should have triggered a responseb | 17 (47.2) | 0 | 16 (28.1) | 0 | 1.00 | ||
| All parameters recorded | 5 (13.9) | 0 | 0.003 | 20.08b | 1.08-375.09b | ||
| Incomplete recording of all parameters | 31 (86.1) | 57 (100) | |||||
For 21 patients in the intervention arm, the MEWS chart had not been used.
aMEWS trigger of 1 = recheck measurement after half an hour and report if no improvement; MEWS trigger of 2 = recheck measurement after five minutes and report immediately if no improvement; MEWS trigger of 3 = critical, report urgently.
bHaldane’s estimator was used for calculating OR (this circumvents zero values in cells by adding one half to each cell).
In the intervention wards nurses responded to 3.4% (three out of 88) MEWS that should have triggered reporting versus 1.1% (one out of 90) in the control wards (chi-squared = 1.07, df = 1, P = 0.30). Overall, nurses in both arms reported a combined total of four out of 178 (2.2%) deranged physiological parameters for four out of 93 (4.3%) patients who needed to be assessed (three in the intervention arm, and one in the control arm). CI, confidence interval; MEWS, modified early warning score; OR, odds ratio; df, the degrees of freedom, or the number of ‘entities’ that are free to vary when a statistical test is applied; this also determines the probability distribution used for the test statistic (see footnote to Table 3).