Skip to main content

Table 3 Coefficients (and associated standard errors) from the propensity-score adjusted difference-in-differences regressions for the effects of pay for performance on healthcare costs among adult patients diagnosed with diabetes before the study period (baseline cohort), by provider’s sex

From: Gendered effects of pay for performance among family physicians for chronic disease care: an economic evaluation in a context of universal health coverage

  (1) (2) (3) (4)
Physician workforce costs Hospital costs for diabetes Hospital costs for comorbid conditions All-cause healthcare costs
Exposed group Control group Exposed group Control group Exposed group Control group Exposed group Control group
1. Female providers
 Pre-P4P 6.266 6.319 0.099 0.192 0.366 0.629 6.485 6.741
 Difference − 0.053** (0.016) − 0.092** (0.018) − 2.63** (0.023) − 0.256** (0.022)
 Post-P4P 6.503 6.339 0.100 0.144 0.411 0.591 6.763 6.754
 Difference 0.165** (0.015) − 0.044** (0.014) − 0.181** (0.023) 0.009 (0.019)
 Diff-in-diff 0.218** (0.024) 0.049* (0.021) 0.082* (0.035) 0.265** (0.031)
2. Male providers
 Pre-P4P 6.229 6.321 0.108 0.180 0.405 0.660 6.462 6.760
 Difference − 0.093** (0.010) − 0.071** (0.010) − 0.255** (0.018) − 0.298** (0.015)
 Post-P4P 6.460 6.400 0.118 0.187 0.444 0.710 6.726 6.837
 Difference 0.059** (0.008) −0.069** (0.009) − 0.266** (0.018) − 0.111** (0.011)
 Diff-in-diff 0.152** (0.011) 0.002 (0.014) −0.011 (0.024) 0.188** (0.018)
  1. Note: **p < 0.01, *p < 0.05. Coefficients (with a bootstrap estimation of robust standard errors in parentheses) calculated using propensity score difference-in-differences estimation. Outcomes are log healthcare costs in inflation-adjusted Canadian dollars. Matching variables include patient’s age group, sex, urban/rural residence, comorbid conditions (hypertension, ischemic heart disease), and physician practice variables (remuneration model, practice size). Baseline cohort includes adults aged 35 and over diagnosed with diabetes before the period of observation (N = 42 896). Models are split by the sex of the patient’s most responsible provider
  2. Source: Linked provincial administrative health datasets