Table 6.

Comparison of models based on verbalised and documented safety-netting advicea

Codes from observing consultation/linked dataVerbalised problems (n= 465)Documented problems (n= 415)
SNA present, %OR (95% CI)P-valueSNA present, %OR (95% CI)P-value
Acute/AoC problem52.82.14 (1.27 to 3.62)0.00430.23.91 (1.78 to 8.61)0.001
Chronic problem36.111.0
First presentation53.10.97 (0.54 to 1.76)0.9235.61.97 (0.93 to 4.20)0.078
Not first44.419.2
Single problem in consultationb61.82.58 (1.51 to 4.43)0.00130.92.11 (1.07 to 4.18)0.032
Multiple problems in consultation39.317.4
Follow-up presentb43.80.69 (0.42 to 1.14)0.1519.10.42 (0.22 to 0.80)0.009
No follow-up51.729.1
GP aged <50 years53.92.88 (1.60 to 5.16)<0.00129.03.77 (1.16 to 12.30)0.028
GP aged ≥50 years32.711.8
Patient age ≥65 years47.01.23 (0.68 to 2.22)0.5022.71.30 (0.61 to 2.74)0.50
Patient age 18–64 years45.922.6
Patient sex: female46.30.84 (0.50 to 1.40)0.5022.71.20 (0.61 to 2.36)0.59
Patient sex: male46.122.6
Patient ethnicity: other54.81.15 (0.47 to 2.83)0.7636.41.18 (0.35 to 3.95)0.79
Patient ethnicity: white45.421.5
Index of Multiple Deprivation quintileWald test P = 0.59Wald test P = 0.57
  • a Data from 274 consultations. Multivariable multilevel mixed-effects modelling to adjust for problems seen by the same GP and multiple problems raised by the same patient. Problems with missing data were excluded.

  • b Codes assessed separately based on model type, for example, if two problems discussed but only one documented, coded as single problem consultation in documentation model. AoC = acute on chronic. CI = confidence interval. OR = odds ratios. SNA = safety-netting advice.