Journal of Substance Abuse Treatment 35 (2008) 462–469

Brief article

The economic costs of substance abuse treatment: Updated estimates andcost bands for program assessment and reimbursement

Michael T. French, (Ph.D.)a,b,c,d,⁎, Ioana Popovici, (Ph.D.)d, Lauren Tapsell, (B.A.)d

aDepartment of Sociology, University of Miami, Coral Gables, FL 33124-2030, USAbDepartment of Economics, University of Miami, Coral Gables, FL 33124-2030, USA

cDepartment of Epidemiology and Public Health, University of Miami, Coral Gables, FL 33124-2030, USAdHealth Economics Research Group, Sociology Research Center, University of Miami, Coral Gables, FL 33146-0719, USA

Received 29 August 2007; received in revised form 21 November 2007; accepted 25 December 2007

Abstract

Federal, state, and local government agencies require current and accurate cost information for publicly funded substance abuse treatmentprograms to guide program assessments and reimbursement decisions. The Center for Substance Abuse Treatment published a list ofmodality-specific cost bands for this purpose in 2002. However, the upper and lower values in these ranges are so wide that they offer littlepractical guidance for funding agencies. Thus, the dual purpose of this investigation was to assemble the most current and comprehensive setof economic cost estimates from the readily available literature and then use these estimates to develop updated modality-specific cost bandsfor more reasonable reimbursement policies. Although cost estimates were scant for some modalities, the recommended cost bands are basedon the best available economic research, and we believe that these new ranges will be more useful to and pertinent for all stakeholders ofpublicly funded substance abuse treatment. © 2008 Elsevier Inc. All rights reserved.

Keywords: Economic cost; Substance abuse treatment; Reimbursement; Policy

1. Introduction

Total spending on substance abuse treatment in the UnitedStates was an estimated US$21 billion in 2003 (Mark et al.,2007). The vast majority (77%) of this spending wasfinanced by public sources, including federal, state, and localgovernments. Some of the most difficult decisions that stategovernment policymakers face concern which substance

The authors are entirely responsible for the research and results reportedin this article, and their position or opinions do not necessarily representthose of the Substance Abuse and Mental Health Services Administration,the National Institute on Drug Abuse, the National Institute on AlcoholAbuse and Alcoholism, or the University of Miami.

⁎ Corresponding author. Department of Sociology, University ofMiami, 5202 University Drive, Merrick Building, Room 121F, PO Box248162, Coral Gables, FL 33124-2030, USA. Tel.: +1 305 284 6039; fax: +1305 284 5310.

E-mail address: mfrench@miami.edu (M.T. French).

0740-5472/08/$ – see front matter © 2008 Elsevier Inc. All rights reserved.doi:10.1016/j.jsat.2007.12.008

abuse treatment programs to fund and where to setreimbursement rates. Similarly, a challenge facing bothfederal and state policymakers is how to efficiently allocatelimited public funds while avoiding either overtreatment orundertreatment of substance-abusing patients. In principle,only programs that meet a minimum set of performancestandards should be funded, and reimbursement rates shouldbe linked to the economic (opportunity) cost of serviceprovision,1 clinical outcomes, and economic benefits. Inreality, performance standards are hard to define and enforce,and reimbursement rates are often based on crude approx-imations of economic costs or historical stipends to providersbased solely on capacity estimates.

1 Economic or opportunity costs do not always equate to the accountingcosts of a treatment program. Opportunity costs include the value of unpaidresources, such as the estimated market value of donated buildings orvoluntary labor. Opportunity cost is the value of the next best alternativeforgone when these resources are used by the program.

463M.T. French et al. / Journal of Substance Abuse Treatment 35 (2008) 462–469

Little guidance exists on how reimbursements should becalculated for treatment programs, especially given differ-ences in performance, client characteristics, setting, location,and resource cost. Even among programs within the samemodality, there are considerable differences in the range andintensity of services provided, types of clients served, andlocation of the program. These factors contribute to thelegitimate variation in treatment costs and make it challen-ging to calculate standardized reimbursement rates. Onepotential reimbursement strategy is to use mean program costper client. However, focusing solely on mean cost, even ifmodality specific, would ignore the diversity of treatmentprograms currently operating in the United States. Apotential consequence of failing to recognize and adjust forthese factors is reduced access to treatment for persons withmultiple disabilities (e.g., severe mental illness; pregnantwomen, women with children, or both; criminal justicereferrals; HIV-infected persons) (Cartwright, in press-b).Similarly, the lack of periodic rate adjustments to reflectincreases in operating costs (i.e., effective rate reductions)could cause the elimination of services to the most difficult-to-treat patients (Hser, Anglin, Grella, Longshore, &Prendergast, 1997).

As an alternative to mean cost per patient, the Center forSubstance Abuse Treatment (CSAT) tried to partiallyaddress this informational need by developing a set ofmodality-specific “cost bands” that government policy-makers could use to guide rate-setting decisions (SubstanceAbuse and Mental Health Services Administration[SAMHSA], 2004). Although cost bands are preferable topoint estimates such as mean cost, numerous limitations arepresent as well. First and most notably, the upper and lowerbounds for each modality are so wide that policymakers havelittle practical assistance on where to set specific reimburse-ment amounts for individual programs, even within thebands. For example, CSAT reports a range of US$1,000–5,000 per patient for outpatient nonmethadone treatment,which translates to a weekly cost of approximately US$19–96 per patient. Second, it is not clear how these ranges weredeveloped, to what unit of service they apply (session, day,week, episode, and so on), and whether the estimates werebased on economic costs, accounting costs, programbudgets, or some other factors. Third, one would presumethat, given these wide ranges, some programs will bereimbursed at rates below the midpoint and others will bereimbursed at rates above the midpoint, but criteria statinghow the reimbursement rate decisions are made are notprovided in the source document. Finally, it is not stated if orhow these ranges change over time as treatment services andresource costs change.

Given the pressing need for current and rigorouseconomic cost information that can assist governmentpolicymakers with their funding decisions, we set out toupdate and improve the substance abuse treatment costbands originally proposed by CSAT by linking themdirectly to existing economic research. A key aim of this

investigation is to develop cost bands based uponstandard economic analysis principles, reflective of thecurrent economic literature, that are expressed in 2006dollars, that can be applied to both a typical week and anaverage episode of treatment, and that are better specified(apply to more existing modalities) than the originalCSAT bands.

To complete these objectives, we compiled data from 110substance abuse treatment programs that had completed theDrug Abuse Treatment Cost Analysis Program (DATCAP)(Bradley, French, & Rachal, 1994; French, Bradley, Calin-gaert, Dennis, & Karuntzos, 1994; French, Dunlap, Zarkin,McGeary, & McLellan, 1997; French & McGeary, 1997;Roebuck, French, & McLellan, 2003; www.DATCAP.com).The DATCAP is a program-level data collection instrumentthat is designed to estimate the costs of a substance abusetreatment program based on standard economic principles. Itcan be used in a variety of treatment settings. By using onlyDATCAP estimates, we ensured a standard method of costdata collection and calculation among all programs includedin the cost bands. Using the weekly and episode costestimates obtained through the DATCAP for each program,we estimated new cost bands for eight treatment modalitiesbased on modality-specific interquartile ranges.

We further discuss the DATCAP instrument and themethod for calculating updated cost bands in the Methodssection. We then present the results of the DATCAP and costband estimation, as well as a sensitivity analysis based onother cost estimation approaches. Finally, we outline thestrengths and limitations of the updated cost band estimates,highlight policy implications, and make recommendationsfor future studies.

2. Methods

The DATCAP was the first standardized approach tocollect resource use data and to estimate the economic costsof substance abuse treatment for individual programs andclients. The DATCAP has undergone several transformationssince the early 1990s to improve clarity, coverage, respondentburden, electronic entry, and precision. (For previousversions of and improvements on the DATCAP, see Dunlap& French, 1998; French et al., 1997; French & McGeary,1997; French, Roebuck, et al., 2002; French, Salomé, &Carney, 2002; McCollister, French, Inciardi, et al., 2003;McCollister, French, Prendergast, et al., 2003; Zavala et al.,2005.) Roebuck et al. (2003) published the first summaryarticle of all DATCAP studies to date, which amounted to 85substance abuse treatment programs across 9 modalities.Since 2002, at least 25 additional programs have completedthe DATCAP. The current editions of the full and briefDATCAP instruments can be found at www.DATCAP.com.

The instrument is designed to collect and organizedetailed information on resources used in service deliveryand their associated dollar cost. Resource categories include

2 A detailed version of Table 1 containing program-specific informationis available upon request from the corresponding author.

464 M.T. French et al. / Journal of Substance Abuse Treatment 35 (2008) 462–469

personnel, supplies and materials, contracted services,buildings and facilities, equipment, and miscellaneousitems. The instrument also collects information on programrevenues and client case flows. Data are collected directlyfrom program officials, with the assistance of a researchertrained to administer the DATCAP. An instruction manual,along with the actual data collection instrument, is alsoprovided to program personnel. Total annual economic costsare based on the data reported for the six resource categories.Estimates for average (per client) weekly and episode costsare calculated from program data on average daily census,average length of stay, and total annual cost.

Because the DATCAP estimates are standardized, basedon established economic cost methods, are more volumi-nous, and cover more modalities than the other approaches(see Sensitivity Analysis section), we chose to base theupdated cost bands on DATCAP estimates only. Our firststep in updating the cost bands was to assemble all of theavailable DATCAP estimates since the instrument's incep-tion. This amounted to 110 programs, which we organizedby the following modalities: methadone maintenance,nonmethadone outpatient, intensive (i.e., day treatment)outpatient, adolescent outpatient, screening and brief inter-vention, drug court, adult residential, adolescent residential,therapeutic community, and in-prison therapeutic commu-nity. Because financial information was collected andreported during different years, all cost estimates wereconverted to constant 2006 dollars using the consumer priceindex (CPI).

After assembling the data, we used a statistical softwareprogram (Stata) to calculate an interquartile range (with thevalues corresponding to the 25th and the 75th percentiles ofthe distribution) for weekly and episode costs within eightmodalities. Interquartile ranges are not reported for “screen-ing and brief intervention” and “adolescent residential” dueto very small sample sizes for these modalities. We chose tobase the cost bands on the interquartile range after exploringthe option of running modality-specific linear regressionmodels and calculating 95% confidence intervals for meanpredicted values of weekly and episode costs. The relativelysmall sample sizes within each modality constrained the useof ordinary least squares regression or any other regressiontechnique to derive confidence intervals (Montenegro,2001). Moreover, although a substantially larger samplesize would improve the efficiency of ordinary least squares,95% confidence intervals are directly linked to the number ofprogram-level cost estimates included in the regressions viathe standard errors of the estimates. For example, if the 95%confidence interval was based on 1,000 programs, it wouldbe virtually a point estimate, with little variability around themean value. For this critical reason, the interquartile range isthe preferred approach for estimating the cost bands givenrelatively small sample sizes. In addition, interquartile rangesare advantageous because the estimates can be updated as thenumber of programs available for analysis increases withineach modality.

3. Results

3.1. DATCAP cost estimates

Table 1 presents summary statistics for the DATCAPestimates, by modality. For each of the 110 programs, weused data on average length of stay (reported in weeks),average daily census, total annual economic cost, weeklyeconomic cost per client, and economic cost per treatmentepisode to calculate summary statistics (mean, standarddeviation, standard error, and median) for each of the 10modalities. Some modalities, such as nonmethadoneoutpatient or adolescent outpatient, are represented byat least 20 unique programs and have fairly tightdistributions for average weekly cost and average episodecost. However, other modalities, such as therapeuticcommunity, are represented by fewer programs and/orhave highly variable program-specific cost estimates.2

The sizeable variation in weekly and episode costs forsome modalities can be explained by the substantialdifferences within modality in types of populationsserved, range and intensity of services provided, typesof substance abuse problems addressed, or location of thetreatment programs.

Besides program and client characteristics, two criticaldeterminants of the variation in mean costs are the averagedaily census and the mean length of stay in treatment.Current studies suggest that average daily census affectsprogram-specific operating costs through economies ofscale. Beaston-Blaakman, Shepard, Horgan, and Ritter(2007) and Duffy, Dunlap, Feder, and Zarkin (2004) foundthat larger outpatient treatment facilities are less costly on aper-admission basis than smaller ones, holding constant theclients' characteristics. Harwood, Kallinis, and Liu (2001a)found economies of scale in the residential sector, althoughthis is partly due to lower intensity treatment delivered bylarger residential care providers. To illustrate this phenom-enon, we provide two examples from our dataset. Theaverage daily census for the nonmethadone outpatientprogram with the largest mean weekly cost (US$385) is 22clients, whereas the average daily census for the programwith the smallest mean weekly cost in this modality (US$32)is 388 clients. Another illustration of this point can be foundin the adult residential modality, where the program with thelargest average weekly cost (US$2,070) has an average dailycensus of 8 clients, whereas the program with the smallestaverage weekly cost (US$213) has an average daily censusof 93 clients. Furthermore, the mean length of stay intreatment has a direct influence on average episode costsbecause more resources are necessary as the duration oftreatment grows longer. This positive correlation is evidentin our dataset.

Table 1Summary DATCAP information from 110 substance abuse treatment programs (in 2006 dollars)

Descriptivestatistics

Average lengthof stay (weeks)

Average dailycensus

Total annualeconomic cost (US$)

Weekly economiccost per client a (US$)

Economic cost pertreatment episode b (US$)

Methadone maintenance (n = 12)M 87 377 1,869,042 104 7,409(SD) (42) (182) (750,828) (36) (4,747)(SE) (12) (52) (216,745) (10) (1,794)Mdn 87 350 1,974,738 101 6,689Nonmethadone outpatient (n = 21)M 18 185 974,762 140 2,325(SD) (9) (174) (811,629) (106) (2,183)(SE) (2) (38) (177,112) (23) (476)Mdn 16 165 781,633 90 1,488Intensive outpatient (n = 9)M 12 31 355,532 370 4,271(SD) (8) (39) (353,060) (246) (5,057)(SE) (3) (13) (117,687) (82) (1,788)Mdn 12 17 193,459 283 1,875Adolescent outpatient (n = 20)M 12 10 74,426 218 2,954(SD) (6) (9) (92,669) (99) (2,678)(SE) (1) (2) (20,721) (22) (599)Mdn 13 8 33,405 181 2,104Screening and brief intervention c (n = 3)M N/A N/A 80,856 N/A 407(SD) N/A N/A (47,345) N/A (195)(SE) N/A N/A (27,335) N/A (113)Mdn N/A N/A 104,300 N/A 303Drug court (n = 9)M 46 205 614,316 93 3,942(SD) (20) (184) (155,626) (49) (2,490)(SE) (7) (61) (51,875) (16) (830)Mdn 45 84 617,503 99 2,688Adult residential (n = 22)M 13 34 1,286,840 789 10,228(SD) (13) (20) (772,887) (362) (11,354)(SE) (3) (4) (164,780) (77) (2,478)Mdn 11 30 1,061,395 774 7,405Adolescent residential (n = 1)1 8 22 1,487,883 1,295 10,640Therapeutic community (n = 5)M 33 152 3,790,828 668 21,404(SD) (22) (265) (5,488,606) (221) (14,125)(SE) (10) (119) (2,454,579) (99) (6,317)Mdn 32 53 1,568,042 642 21,251In-prison therapeutic community d (n = 8)M 28 265 1,232,842 63 1,747(SD) (11) (288) (1,806,580) (13) (1,078)(SE) (4) (102) (638,723) (5) (407)Mdn 27 119 365,134 61 1,536

Notes. The table includes data from Fleming et al. (2000), French et al. (in press), Knealing, Roebuck, Wong, and Silverman (in press), Kunz, French, andBazargan-Hejazi (2004), Mundt, French, Roebuck, Baier Manwell, and Lawton Barry (2005), Roebuck et al. (2003), Zavala et al. (2005), and numerousunpublished DATCAP studies.All costs are inflated to 2006 dollars.Some numbers may not add, divide, or multiply exactly due to rounding.N/A, not available.

a Weekly economic cost per client = total annual economic cost / average daily census / 52.14 weeks.b Economic cost per treatment episode = weekly economic cost per client × average length of stay (weeks).c We conducted extensive searches of the relevant literature on screening and brief intervention, and the results of only these three studies provided the

information necessary in the table and subsequent analyses. Furthermore, as the DATCAP was not originally designed to evaluate costs of screening and briefinterventions, Fleming et al. (2000) did not directly use the DATCAP, but a similar method, to estimate the cost of their intervention.

d Reported costs for in-prison therapeutic communities are incremental (i.e., in addition to standard incarceration).

465M.T. French et al. / Journal of Substance Abuse Treatment 35 (2008) 462–469

466 M.T. French et al. / Journal of Substance Abuse Treatment 35 (2008) 462–469

3.2. Cost band estimates

Table 2 presents the recommended cost bands for eachmodality. Compared to the CSAT cost bands discussedearlier, these ranges are generally tighter. CSAT publishedcost bands for an average episode of treatment for three ofthe eight modalities covered by our cost band estimates.For nonmethadone outpatient treatment, our proposed costband (US$1,132–2,099) is much narrower than CSAT'scost band (US$1,000–5,000). Alternatively, the extremevariation in average length of stay for methadonemaintenance (16–150 weeks) and adult residential (3–50weeks) programs in our dataset makes the proposed costbands for these modalities (see Table 2) wider than thosepresented by CSAT (US$1,500–8,000 for methadonemaintenance and US$3,000–10,000 for adult residential).

Of course, some currently operating programs in eachmodality may have average weekly or episode costs that fallconsiderably below or rise well above the proposed ranges.Much of the variation in mean episode costs will be due toaverage length of stay in treatment, which can be highlyvariable across programs. We would view these “out-of-range” programs as true outliers that may require a differentset of criteria for reimbursement decisions, which should notnecessarily be factored into the prevailing cost bands. Stateddifferently, the updated research-based cost bands presentedin Table 2 are intended to characterize the majority ofprograms in each modality that are more typical in terms ofservice delivery, average length of stay, and costs. Recogniz-ing that unusual programs exist, other program features mustbe considered to derive alternative reimbursement guidelinesfor these entities. For example, special populations (e.g.,dually diagnosed patients, women with children) areexpensive to treat, as costs include wrap-around servicesthat supplement the core services provided to the generalpopulation. These wrap-around services are often highlycorrelated with treatment retention and outcomes. Therefore,reimbursement rates for programs treating special popula-tions should take into account the additional costs incurred asa result of these specialized services. Moreover, the chronic

Table 2Proposed weekly and episode cost bands

ModalityWeekly cost(US$)

Episode cost(US$)

Methadone maintenance 87–112 4,277–13,395Nonmethadone outpatient 74–221 1,132–2,099Intensive (i.e., day treatment) outpatient 243–598 1,384–5,780Adolescent outpatient 139–281 1,517–3,237Drug court 34–146 2,486–4,888Adult residential 607–918 2,907–11,260Therapeutic community 569–708 14,818–32,361In-prison therapeutic community 55–71 1,249–2,112

Notes. The cost bands correspond to the interquartile range for eachmodality (25th–75th percentiles). Estimates are not reported for “screeningand brief intervention” and “adolescent residential” due to very smallsample sizes for these modalities.

and relapsing nature of addiction points to the need forextended care or some form of continuing care following theinitial phase of treatment. Emerging clinical literatureprovides support for the effectiveness of extended orcontinuing care, but the optimum length of stay in treatmentis still the subject of intense debate (McKay, 2005, 2007).The large variation in the mean length of stay amongprograms within the same modality is the main factoraffecting the width of cost bands for average episode cost.

3.3. Sensitivity analysis

A natural sensitivity or legitimacy test for the proposedcost bands is how they compare to the published costestimates using data collection methods other than theDATCAP. Several other substance abuse treatment costestimation approaches have been proposed since the launchof the first edition of the DATCAP in 1994. Among these arethe Substance Abuse Services Cost Analysis Program(Zarkin, Dunlap, & Homsi, 2004), the Alcohol and DrugServices Study (SAMHSA, 2003), Capital Consulting's(1998) accounting approach, and the unit service costs andaccounting method (Anderson, Bowland, Cartwright, &Bassin, 1998). Besides these instruments, several publishedstudies use alternative methods to estimate daily, weekly,episode, or total annual costs of treatment. Extensiveliterature reviews were conducted to locate all of the possiblestudies that estimate the costs of addiction treatment. Wesearched Web of Science and Google Scholar using thekeywords “substance abuse treatment” or “addiction treat-ment” and “cost” to identify studies that provide addictiontreatment cost estimates. In selecting cost estimates fromother sources, we included studies that were readily availablein the literature, that were based on standard economic oraccounting principles, and that reported findings that wereidentical or at least similar to the DATCAP cost categories.We excluded roughly 10 studies that used budgets rather thanexpenditures, reimbursement rates, or billing records toestimate costs, as they were not directly comparable to theDATCAP cost estimates. We identified 13 studies that fit ourinclusion criteria (Alterman et al., 1994; Anderson et al.,1998; Avants et al., 1999; Barnett & Swindle, 1997;Burgdorf, Layne, Roberts, Miles, & Herrell, 2004; Ettneret al., 2006; Hartz et al., 1999; Harwood et al., 2001a, 2001b;Mojtabai & Zivin, 2003; SAMHSA, 2003; Weisner et al.,2000; Zarkin et al., 2004). The majority of these studiesestimate economic costs per client per day, episode, month,or year of treatment. These economic cost estimates includethe opportunity cost of the resources used, such as the valueof volunteer labor or donated buildings, making themsimilar and comparable to the DATCAP estimates. Whenthe study did not provide average weekly costs, weestimated these based on the data and estimates provided.For example, the weekly cost per client was calculatedby multiplying the mean cost per client day by 7 days/week,by dividing the average cost per month by 4.348, or by

467M.T. French et al. / Journal of Substance Abuse Treatment 35 (2008) 462–469

dividing the mean cost per episode by the average numberof days in treatment and multiplying the result by 7 days/week. All cost estimates were converted to 2006 dollarsusing the CPI. We compiled these cost estimates into a tablesimilar to Table 1 that is available upon request. We thencompared them against the cost bands derived from theDATCAP data for each modality.

The 13 studies included in the sensitivity analysisestimate costs for 849 programs across 5 modalities: 244methadone maintenance programs, 280 nonmethadone out-patient programs, 12 intensive outpatient programs, 305residential programs, and 8 methadone detoxificationprograms. Some of the estimates are reported for individualprograms, whereas others are averages of multiple programswithin the same modality. Ninety percent of the modality-specific cost estimates fall within the proposed episode costbands in Table 2 or very close to the lower or upper bounds.Eighty-eight percent of the weekly cost estimates fall withinor very close to the proposed weekly cost bands. The episodecost estimates that fall outside the proposed cost bands(corresponding to 27 methadone, 9 outpatient, 1 intensiveoutpatient, and 39 residential programs) can be attributed toatypical values for the average length of stay when comparedto the average length of stay reported by the programscompleting the DATCAP. We suspect that a few weekly costestimates fall outside the proposed cost bands (correspond-ing to 1 methadone, 1 intensive outpatient, and 90 residentialprograms) due to the considerable diversity in the intensity orrange of treatment services provided by programs across thecountry. With the additional programs included in thissensitivity analysis, we were able to consider cost data fornearly 1,000 substance abuse treatment programs across theUnited States. Given that the vast majority of program costsadhere closely to the updated cost bands presented inTable 2, this sensitivity analysis bolsters our confidence inthe DATCAP-based cost bands.

4. Discussion

The substance abuse treatment cost bands originallyproposed by CSAT need to be updated and improved withevidence from standardized and scientific studies of treat-ment services, duration, and economic costs if they are tobecome meaningful to and constructive for policymakers andprogram evaluators (see French et al., 1994). The presentanalysis incorporated DATCAP-based economic cost esti-mates from a variety of sources to derive research-driven costbands for most of the major substance abuse treatmentmodalities. The proposed cost bands cover more modalities,are portrayed in 2006 dollars, apply to both a week oftreatment and an episode of treatment, and can easily belinked to the data in Table 1.

We must caution that any reimbursement rates derivedfrom the cost bands should be applied to individual programscarefully, as the cost bands alone are not intended for use as

efficiency measures. Programs that operate efficiently mayfall above or below certain cost bands due to the variations inservices provided, clients served, program location, andother factors. Conversely, an inefficient program may stillfall within the estimated cost bands due, for example, tolower resource costs in their area of operation. A separate setof performance or evaluation measures should be used inaddition to the cost bands when undertaking a full economicevaluation of substance abuse treatment programs. Spaceconstraints do not permit us to fully elaborate on the methodsused for more advanced approaches, such as cost-effective-ness analysis, cost–utility analysis, and benefit–cost analy-sis. These techniques have been presented thoroughly in anumber of recent articles in the substance abuse literature(e.g., Cartwright, 2000; Cartwright, in press-a, in press-b;French & Drummond, 2005; Zavala et al., 2005). Interestedreaders can consult these articles to obtain a completeperspective on the approaches used to estimate costs,outcomes, and economic benefits of substance abusetreatment services.

4.1. Limitations

Despite the improvements made by this study, there arestill limitations to the updated cost bands. One limitationpertains to the cost per treatment episode. As mentionedpreviously, there is large variation in the average length ofstay within most treatment modalities. In addition, there is noconsensus in the clinical literature on the appropriate lengthof treatment episodes for all types of clients; episode lengthis highly variable based on individual client needs andprogram characteristics (Hser et al., 1997). For example,among methadone maintenance treatment programs in oursample, the average length of stay (which constitutes oneepisode of treatment) ranges from 16 to 150 weeks. Thesevariations make comparisons of episode costs acrossprograms tenuous. Although variations in the intensity,range of treatment services offered, patients served, andlocation of the program result in differences in weekly costsas well, we believe that the weekly cost bands are morestable and reliable as a unit of comparison than the episodecost bands.

A challenge related to cost bands as a tool forreimbursement decisions is the lack of corresponding dataon program and client characteristics. For example, specialpopulations, such as patients with multiple co-occurringconditions (e.g., clients with substance abuse and mentaldisorders), can be very expensive to treat. The proposed costbands do not take into account the comprehensive servicesthat should be provided to these patients. On the other hand,a program that primarily treats functional patients withsupport systems and who are still employed may generateactual costs that fall below the proposed bands. Unfortu-nately, we do not have the necessary data to recommendadjustments to the cost bands based on program and clientcharacteristics, and especially clinical outcomes. We cannot

468 M.T. French et al. / Journal of Substance Abuse Treatment 35 (2008) 462–469

determine, at this point, whether programs with higheroperating costs provide better or more effective treatment.Indeed, the higher costs may be more reflective of theservices needed to address populations with greater problemseverity, and hence poorer prognoses for improvement.Finally, besides applying the CPI or some other index toinflate the current estimates to future years, there is nostructured way to update the cost bands based on organiza-tional, service delivery, client, or resource changes.

4.2. Conclusion

This study improves upon and updates the cost bandsoriginally developed by CSAT by basing the values on acomprehensive set of economic cost estimates derivedthrough a standardized, peer-reviewed, and well-usedinstrument. However, given the limitations that still existwith the new cost bands, we encourage CSAT, the NationalInstitute on Drug Abuse, the National Institute on AlcoholAbuse and Alcoholism, and other interested stakeholders toembark on a comprehensive study of substance abusetreatment costs that can be used to develop, among otherthings, a defensible set of modality-specific cost bands thatcan be adjusted for program and client characteristics andoutcomes, and that can be updated periodically to reflectcurrent resource costs and trends.

One way for CSAT and other stakeholders to undertakethe suggested study is to use an augmented version of theDATCAP to collect cost and key clinical information on anationally representative sample of programs within eachtreatment modality. By expanding the instrument to includeadditional information on a variety of program and clientcharacteristics (e.g., urban, suburban, or rural setting;program size; ownership; service provision; clients withspecial needs), more rigorous models and estimates can bedeveloped. This enhancement would allow researchers tomake adjustments in the cost bands for changes in programorganization, service delivery, clients, or resources, all ofwhich can affect economic costs.

We feel confident that these updated cost bands are asignificant improvement upon the original ranges publishedby CSAT. These estimates can be used to formulate futurereimbursement decisions for existing programs, to estimatethe costs of treatment expansion, and to shape the foundationof more advanced economic evaluations. However, ouranalysis indicates that a comprehensive study as describedabove is necessary to make any further improvements in thecost bands. We hope that this article serves as a motivatorand guide for CSAT and other interested stakeholders toundertake such a study of substance abuse treatment costs inthe near future.

Acknowledgments

Financial assistance for this study was provided by theSAMHSA (contract GS-10G-F-0207L; task order no. 283-

05-0140), the National Institute on Drug Abuse (R01DA018645), and the National Institute on Alcohol Abuseand Alcoholism (R01 AA13167). We gratefully acknowl-edge Sue Becker, Michael Dennis, Sarah Duffy, LauraDunlap, Rick Harwood, Alisa Male, Kathryn McCollister,Thomas McLellan, Steven Milas, Christopher Roebuck,Donald Shepard, Gary Zarkin, and Silvana Zavala fortechnical suggestions on earlier versions of the article; SeanLi and Robin Prize for research assistance; and CarmenMartinez, Elizabeth Walter, Steven Milas, and Robin Prizefor editorial assistance.

References

Alterman, A. I., O'Brien, C. P., McLellan, A. T., August, D. S., Snider, E. C.,Droba, M., et al. (1994). Effectiveness and costs of inpatient versus dayhospital cocaine rehabilitation. The Journal of Nervous and MentalDisease, 182, 157−163.

Anderson, D. W., Bowland, B. J., Cartwright, W. S., & Bassin, G. (1998).Service-level costing of drug abuse treatment. Journal of SubstanceAbuse Treatment, 15, 201−211.

Avants, S. K., Margolin, A., Sindelar, J. L., Rounsaville, B. J., Schottenfeld,R., Stine, S., et al. (1999). Day treatment versus enhanced standardmethadone services for opioid-dependent patients: A comparison ofclinical efficacy and cost. American Journal of Psychiatry, 156, 27−33.

Barnett, P. G., & Swindle, R. W. (1997). Cost-effectiveness of inpatientsubstance abuse treatment. Health Services Research, 32, 615−629.

Beaston-Blaakman, A., Shepard, D., Horgan, C., & Ritter, G. (2007).Organizational and client determinants of cost in outpatient substanceabuse treatment. The Journal of Mental Health Policy and Economics,10, 3−13.

Bradley, C. J., French, M. T., & Rachal, J. V. (1994). Financing and cost ofstandard and enhanced methadone treatment. Journal of SubstanceAbuse Treatment, 11, 433−442.

Burgdorf, K., Layne, M., Roberts, T., Miles, D., & Herrell, J. M. (2004).Economic costs of residential substance abuse treatment for pregnantand parenting women and their children. Evaluation and ProgramPlanning, 27, 233−240.

Capital Consulting Corporation (CCC). (1998). The Lewin Group andCaliber Associates. Measuring the cost of substance abuse treatmentservices: An overview. Report for the Center for Substance AbuseTreatment. Substance Abuse andMental Health Services Administration.

Cartwright, W. S. (2000). Cost–benefit analysis of drug treatment services:Review of the literature. The Journal of Mental Health Policy andEconomics, 3, 11−26.

Cartwright, W. S. (in press-a). A critical review of accounting and economicmethods for estimating the costs of addiction treatment. Journal ofSubstance Abuse Treatment.

Cartwright, W. S. (in press-b). Economic costs of drug abuse: Financial,cost-of-illness, and services. Journal of Substance Abuse Treatment.

Duffy, S. Q., Dunlap, L. J., Feder, M., & Zarkin, G. A. (2004). A hybrid costfunction for outpatient nonmethadone substance abuse treatmentfacilities. In C. L. Council (Ed.), Health services utilization byindividuals with substance abuse and mental disorders (pp. 133−153).Rockville, MD: Office of Applied Studies, Substance Abuse and MentalHealth Services Administration [DHHS Publication No. SMA 04-3949,Analytic Series A-25].

Dunlap, L. J., & French, M. T. (1998). A comparison of two methods forestimating the costs of drug abuse treatment. Journal of Maintenance inthe Addictions, 1, 29−44.

Ettner, S. L., Huang, D., Evans, E., Ash, D. R., Hardy, M., Jourabchi, M.,et al. (2006). Benefit–cost in the California Treatment Outcome Project:Does substance abuse treatment “pay for itself”? Health ServicesResearch, 41, 192−213.

469M.T. French et al. / Journal of Substance Abuse Treatment 35 (2008) 462–469

Fleming, M. F., Mundt, M. P., French, M. T., Manwell, L. B., Stauffacher,E. A., & Lawton Barry, K. (2000). Benefit–cost analysis of briefphysician advice with problem drinkers in primary care settings.Medical Care, 38, 7−18.

French,M. T., Bradley, C. J., Calingaert, B., Dennis,M. L., &Karuntzos, G. T.(1994). Cost analysis of training and employment services in methadonetreatment. Evaluation and Program Planning, 17, 107−120.

French, M. T., & Drummond, M. (2005). A research agenda for economicevaluation of substance abuse services. Journal of Substance AbuseTreatment, 29, 125−137.

French,M.T.,Dunlap, L. J., Zarkin,G.A.,McGeary, K.A.,&McLellan, A. T.(1997). A structured instrument for estimating the economic cost of drugabuse treatment: The Drug Abuse Treatment Cost Analysis Program(DATCAP). Journal of Substance Abuse Treatment, 14, 445−455.

French, M. T., & McGeary, K. A. (1997). Estimating the economic cost ofsubstance abuse treatment. Health Economics, 6, 539−544.

French, M. T., Roebuck, M. C., Dennis, M. L., Diamond, G., Godley, S.,Tims, F., et al. (2002). The economic cost of outpatient marijuanatreatment for adolescents: Findings from a multisite field experiment.Addiction, 97(Suppl 1), 84−97.

French, M. T., Salomé, H. J., & Carney, M. (2002). Using the DATCAPand ASI to estimate the costs and benefits of residential addictiontreatment in the state of Washington. Social Science and Medicine, 55,2267−2282.

French, M. T., Zavala, S. K., McCollister, K. E., Waldron, H. B., Turner,C. W., & Ozechowski, T. J. (in press). Cost-effectiveness analysis offour interventions for adolescents with a substance use disorder.Journal of Substance Abuse Treatment.

Hartz, D. T., Meek, P., Piotrowski, N. A., Tusel, D. J., Henke, C. J.,Delucchi, K., et al. (1999). A cost-effectiveness and cost–benefitanalysis of contingency contracting-enhanced methadone detoxifica-tion treatment. American Journal of Drug and Alcohol Abuse, 25,207−218.

Harwood, H. J., Kallinis, S., & Liu, C. (2001). Do larger residential servicedelivery units have lower costs? Report prepared by the NationalEvaluation Data Services (NEDS) under Caliber/NEDS contract no.270-97-7016 supported by the Center for Substance Abuse Treatment,Department of Health and Human Services. Falls Church, VA: The LewinGroup.

Harwood, H. J., Kallinis, S., & Liu, C. (2001). The cost and components ofsubstance abuse treatment. Report prepared by National Evaluation DataServices (NEDS) under Caliber/NEDS contract no. 270-97-7016supported by the Center for Substance Abuse Treatment, Departmentof Health and Human Services. Falls Church, VA: The Lewin Group.

Hser, Y., Anglin, M. D., Grella, C., Longshore, D., & Prendergast, M. L.(1997). Drug treatment careers: A conceptual framework and existingresearch findings. Journal of Substance Abuse Treatment, 14,543−558.

Knealing, T. W., Roebuck, M. C., Wong, C. J., & Silverman, K. (in press).Economic cost of the therapeutic workplace intervention added tomethadone maintenance. Journal of Substance Abuse Treatment.

Kunz, F. M., French, M. T., & Bazargan-Hejazi, S. (2004). Cost-effectiveness analysis of a brief intervention delivered to problem

drinkers presenting at an inner-city hospital emergency department. TheJournal of Studies on Alcohol, 65, 363−370.

Mark, T. L., Levit, K. R., Coffey, R. M., McKusick, D. R., Harwood, H. J.,King, E. C., et al. (2007). National expenditures for mental healthservices and substance abuse treatment, 1993–2003. Rockville, MD:Substance Abuse and Mental Health Services Administration [DHHSPublication No. SMA 07-4227].

McCollister, K. E., French, M. T., Inciardi, J. A., Butzin, C. A., Martin, S. S.,& Hooper, R. M. (2003). Post-release substance abuse treatment forcriminal offenders: A cost-effectiveness analysis. Journal of Quantita-tive Criminology, 19, 389−407.

McCollister, K. E., French, M. T., Prendergast, M., Wexler, H., Sacks, S., &Hall, E. (2003). Is in-prison treatment enough? A cost-effectivenessanalysis of prison-based treatment and aftercare services for substance-abusing offenders. Law and Policy, 25, 63−82.

McKay, J. R. (2005). Is there a case for extended interventions for alcoholand drug use disorders? Addiction, 100, 1594−1610.

McKay, J. R. (2007). Continuing care research: What we've learnedand where we're going. Unpublished manuscript, University ofPennsylvania.

Mojtabai, R., & Zivin, J. G. (2003). Effectiveness and cost-effectiveness offour treatment modalities for substance disorders: A propensity scoreanalysis. Health Services Research, 38, 233−259.

Montenegro, A. (2001). On sample size and precision in ordinary leastsquares. Journal of Applied Statistics, 28, 603−605.

Mundt, M. P., French, M. T., Roebuck, M. C., Baier Manwell, L., & LawtonBarry, K. (2005). Brief physician advice for problem drinking amongolder adults: An economic analysis of costs and benefits. Journal ofStudies on Alcohol, 66, 389−394.

Roebuck, M. C., French, M. T., & McLellan, A. T. (2003). DATStats:Summary results from 85 completed Drug Abuse Treatment CostAnalysis Programs (DATCAPs). Journal of Substance Abuse Treatment,25, 51−57.

Substance Abuse and Mental Health Services Administration (SAMHSA).(2003). The ADSS Cost Study: Costs of substance abuse treatment in thespecialty sector. Rockville, MD: Author [DHHS Publication No. SMA03-3762, Analytic Series A-20].

Substance Abuse and Mental Health Services Administration (SAMHSA).(2004). Access to Recovery (ATR) grants (TI 04-009). Initialannouncement. Retrieved on June 2007 from: http://atr.samhsa.gov/downloads/ti04009_ATR.doc.

Weisner, C., Mertens, J., Parthasarathy, S., Moore, C., Hunkeler, E. M., Hu,T., et al. (2000). The outcome and cost of alcohol and drug treatment inan HMO: Day hospital versus traditional outpatient regimens. HealthServices Research, 35, 791−812.

Zarkin, G. A., Dunlap, L. J., & Homsi, G. (2004). The Substance AbuseServices Cost Analysis Program (SASCAP): A new method forestimating drug treatment services costs. Evaluation and ProgramPlanning, 27, 35−43.

Zavala, S. K., French, M. T., Henderson, C. E., Alberga, L., Rowe, C., &Liddle, H. A. (2005). Guidelines and challenges for estimating theeconomic costs and benefits of adolescent substance abuse treatments.Journal of Substance Abuse Treatment, 29, 191−205.