Table 1.

Monte Carlo Simulation Results to Test the Robustness of an Asthma Quality Indicator^a to Systematic Data Loss

	Baseline Sample		Baseline Sample with Redundant Controller Medication Fills Removed		Baseline Sample with Half of Controller Medication Fills Removed
Loss of All Data for (x) Months		QI Rate	Percent Deterioration in QI rate	QI Rate	Percent Deterioration in QI rate	QI Rate	Percent Deterioration in QI rate
Number	%	Mean (SE)	%	Mean (SE)	%	Mean (SE)	%
0	0.0	.59	0.0	.59	0.0	.29	0.0
		(.00)		(.00)		(.00)
1	16.66	.58	1.7	.53	10.2	.29	0.0
		(.00)		(.07)		(.00)
2	33.33	.58	1.7	.46	22.0	.28	3.4
		(.01)		(.10)		(.00)
3	50.00	.56	5.1	.39	33.9	.27	6.9
		(.01)		(.11)		(.01)
4	66.66	.53	10.2	.30	49.2	.25	13.8
		(.02)		(.12)		(.01)
5	83.33	.46	22.0	.21	64.4	.22	24.2
		(.04)		(.11)		(.02)

^aThe asthma quality indicator is defined as the proportion of children with asthma filling two or more rescue medications who also filled at least one controller medication during a six-month observation period based on fee-for-service claims.

Source: Maryland Medicaid fee-for-service claims data June 1996 and December 1997.