Diabetes treatment and control: the effect of public health insurance for the poor in Mexico
Sandra G Sosa-Rubí, Omar Galárraga & Ruy López-Ridaura
Volume 87, Number 7, July 2009, 512-519
Table 3. Average effect of enrolment in the Seguro Popular in terms of diabetes treatment and control, Mexico, 2006a
Outcome variable | Mean |
Unmatched |
Matched |
||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Seguro Popular | Uninsured | Unmatched differenceb | t-value | P-value | Matched ATTc (95% CI) | t-value | P-value | ||||||||
Health care access outcome | |||||||||||||||
Regular use of any blood glucose control testd | 0.809 | 0.754 | 0.055 | 2.300 | *** | 0.095 (0.020‑0.161) | 2.626 | *** | |||||||
No. of insulin users | 0.0415 | 0.0479 | –0.0064 | –0.504 | NS | 0.001 (–0.032–0.034) | 0.066 | NS | |||||||
No. of insulin shots per week among insulin users | 12.6 | 8.99 | 3.600 | 2.922 | *** | 3.133 (0.04–6.22) | 1.988 | ** | |||||||
No. of blood sugar control tests per month | 1.02 | 0.87 | 0.144 | 0.825 | NS | 0.142 (0.024–0.260) | 2.371 | *** | |||||||
Four or more physician visits per year | 0.918 | 0.811 | 0.107 | 4.795 | *** | 0.093 (0.032–0.154) | 3.019 | *** | |||||||
Biological outcome | |||||||||||||||
Proportion of diabetics with appropriate glucose controle | 0.089 | 0.074 | 0.015 | 0.624 | NS | 0.056 (0.009–0.103) | 2.316 | *** | |||||||
Proportion of diabetics with very poor glucose controlf | 0.367 | 0.462 | –0.095 | –2.122 | *** | –0.175 (–0.285–0.065) | –3.159 | *** |
**P < 0.05; ***P < 0.01, using t tests, of mean differences between treatment and control (unmatched and matched) groups.ATT, average treatment effect on the treated; CI, confidence interval; NS, not significant.
a The data are not weighted.
b This is the raw difference between unmatched samples, or the difference between the mean observed in Seguro Popular enrollees and the mean observed in the uninsured population.
c The ATT is based on the differences in mean outcome for propensity-score-matched individuals using nearest neighbour matching.
d Tests may include urine or capillary blood reactive strips, blood glucose, or glycosylated haemoglobin (HbA1c).
e Appropriate glucose control (HbA1c ≤ 7%) compared with poor glucose control (HbA1c > 7%).
f Very poor glucose control (HbA1c > 12%) compared with normal and poor glucose control (HbA1c ≤ 12%).