Dt. Hartz et al., A cost-effectiveness and cost-benefit analysis of contingency contracting-enhanced methadone detoxification treatment, AM J DRUG A, 25(2), 1999, pp. 207-218
We examined treatment costs in an ongoing study in which 102 opioid-addicte
d patients had been randomly assigned to either 180-day methadone detoxific
ation or the same treatment enhanced with contingency contracting. In the l
atter condition, study participants received regular reinforcers contingent
on negative urine toxicology screens and breath analyses for a range of dr
ugs and alcohol. Both conditions involved psychosocial treatment, and all p
articipants were stabilized to a daily methadone dose of approximately 80 m
g during the first 4 months, followed by a 2-month taper. Individuals parti
cipating in the enhanced condition were more likely to provide continuously
drug-free urine samples and alcohol-free breath samples during the final m
onth of treatment than were participants in the control condition. Cost of
treatment was calculated individually for each participant based on actual
services received. First, unit cost for each service was determined, includ
ing adjusted staff salaries for direct treatment and opportunity cost of fa
cilities utilized during service delivery. Next, we valued each patient's u
se of services during the first 120 days of the study and then added the co
st of methadone, laboratory work, and contingent reinforcers. A subsample (
n = 45) also provided data on health care utilization during treatment, whi
ch we valued using standard Medicare unit costs. The marginal cost of enhan
cing the standard treatment with contingency contracting was approximately
8%. An incremental cost of $17.27 produced an additional 1% increase in the
number of participants providing continuously substance-free urine and bre
ath samples during month 4 of the study. For every additional dollar spent
on treatment. a $4.87 health care cost offset was realized; however. this d
ifference was statistically insignificant due to extreme variances and smal
l subsample size.