Skip to main content

Table 3 Logistic regression analysis for optimal referral pathway-all participants

From: Using routine referral data for patients with knee and hip pain to improve access to specialist care

VariablesSub optimal (n = 209)Optimal (n = 419)Final multivariate regression (n = 386)
number%number%Odds ratio95% confidence intervalp-value
BMI score13632.3 (8.8)25730.0 (6.7)1.00.9, 1.00.004
Diagnostic procedure05232.310967.7Ref0.001
110737.317762.30.50.3, 0.9
2+3325.69674.41.40.7, 2.8
Disease or syndromeNo8740.113059.9Ref0.017
Yes10529.425270.61.81.1, 2.8
Pharmacologic substanceNo16335.329964.7Ref0.039
Yes2925.98374.11.81.0, 3.3
  1. Variables not included in the final multivariate model as they were not significant at 10% level: age, gender, smoking status, welsh index of multiple deprivation (median and quintiles), geographical area, joint type, BMI categories, Charleston co-morbidity index (median and score) number of co-morbidities, medication history, number of semantic types, referral letter length, body related concept, daily or recreational activity, finding, functional concept, health and care activity, injury or poisoning, occupational activity, physiologic function, sign or symptom, tissue, therapeutic or preventative procedure