• 
• 

Three Part Question

In [ED patients presenting with alcohol-related events] does [Brief Intervention when compared with standard care] [reduce cost/ prove more cost-effective?]

Clinical Scenario

Your ED has recently instituted an Alcohol Health Service comprising two designated Alcohol Health Workers who administer brief psychotherapeutic interventions to children and adults presenting with alcohol-related events. You wish to find out whether such intervention is cost-effective...

Search Strategy

MEDLINE 1950 to July week 1 2007
EMBASE 1980 to 2007 week 28:
PsycINFO 1967 to July week 2 2007
CINAHL 1982 to July week 1 2007
The Cochrane Library Issue 2, 2007
MEDLINE:
{[(alcohol$.mp. OR ethanol.mp. OR exp Ethanol/ OR booze$.mp. OR exp Alcohol Drinking/) AND (addict$.mp. OR hazard$.mp. OR problem$.mp. OR binge$.mp. OR abuse$.mp. OR misuse$.mp OR dependence.mp.)] OR [alcoholism.mp. OR exp Alcoholism/ OR exp Alcoholic Intoxication/]}
AND [brief intervention$.mp. OR (brief.mp. AND intervention$.mp.) OR exp Psychotherapy, Brief/ OR exp Counseling/ OR counsel$.mp. OR exp Health Personnel/ OR health worker.mp.]
AND [exp Emergency Service, Hospital/ OR emergency department$.mp. OR (accident and emergency).mp. OR exp Trauma Centers/]
LIMIT to [humans AND English language]
EMBASE:
{[(alcohol$.mp. OR exp ALCOHOL/ OR ethanol.mp. OR booze$.mp. OR exp Alcohol Consumption/) AND (addict$.mp. OR hazard$.mp. OR problem$.mp. OR binge$.mp. OR abuse$.mp. OR misuse$.mp OR dependence.mp.)] OR [alcoholism.mp. OR exp ALCOHOLISM/ OR exp Drug Dependence/ OR exp Drug Misuse/ OR exp Alcohol Abuse/ OR exp ALCOHOL INTOXICATION/]}
AND [brief intervention$.mp. OR (brief.mp. AND intervention$.mp.) OR exp PSYCHOTHERAPY/ OR exp COUNSELING/ OR counsel$.mp. OR exp Health Care Personnel/ OR health worker.mp.]
AND [exp Emergency Medicine/ OR exp Emergency Health Service/ OR emergency department$.mp. OR exp emergency ward/ OR (accident and emergency).mp. OR trauma center.mp.]
LIMIT to [humans AND English language]
PsycINFO:
{[(alcohol$.mp. OR ethanol.mp. OR exp ETHANOL/ OR booze$.mp.) AND (addict$.mp. OR hazard$.mp. OR problem$.mp. OR binge$.mp. OR abuse$.mp. OR misuse$.mp)] OR [alcoholism.mp. OR exp ALCOHOLISM/ OR exp Alcohol Drinking Patterns/ OR exp Drug Dependency/ OR exp ALCOHOL ABUSE/ OR exp Alcohol intoxication/]}
AND [brief intervention$.mp. OR (brief.mp. AND (exp INTERVENTION/ OR intervention$.mp.)) OR exp Brief Psychotherapy/ OR exp Counseling/ OR counsel$.mp. OR exp Health Personnel/ OR health worker.mp.]
AND [exp Emergency Services/ OR emergency department$.mp. OR (accident and emergency).mp. OR trauma center$]
LIMIT to [humans AND English language]
CINAHL:
{[(alcohol$.mp. OR ethanol.mp. OR exp Alcohol, Ethyl/ OR booze$.mp. OR exp Alcohol Drinking/) AND (addict$.mp. OR hazard$.mp. OR problem$.mp. OR binge$.mp. OR abuse$.mp. OR misuse$.mp OR dependence.mp.)] OR [alcoholism.mp. OR exp ALCOHOLISM/ OR exp Alcohol abuse/ OR exp Alcoholic Intoxication/]}
AND [brief intervention$.mp. OR (brief.mp. AND intervention$.mp.) OR exp PSYCHOTHERAPY/ OR exp COUNSELING/ OR counsel$.mp. OR exp Health Personnel/ OR health worker.mp.]
AND [exp Emergency Service/ OR emergency department$.mp. OR (accident and emergency).mp. OR exp Trauma Centers/]
LIMIT to [humans AND English language]
The Cochrane Library:
[Alcohol Drinking/ OR alcohol*.mp. OR problem drinking.mp.] AND [Psychotherapy, Brief/ OR brief intervention.mp.] AND [Emergency Service, Hospital/ OR Emergency Medical Services/ OR emergency department*.mp.]

Search Outcome

590 articles were found in all databases, of which 3 were relevant and of sufficient quality for inclusion

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Barrett B et al. 2006 UK	599 ED patients aged >=18 years screening PAT (Paddington Alcohol Test) positive; consenting; English-speaking; resident within Greater London; no contact with alcohol services; not requesting help with alcohol problems. 312 randomised to control, 287 to brief intervention Control - standard care: health information leaflet 'Think about drink' including list of national help-lines and local alcohol support agencies Intervention - standard care plus appointment with alcohol health worker (mental health nurse): 30mins; assessment and discussion of current and previous drinking; help with resolving ambivalence regarding drinking; feedback about safe amounts of drinking and suggesting range of strategies to reduce alcohol consumption Usable sample of 159 control group and 131 intervention group with follow-up data available at 6 and 12 months	Cost-effectiveness analysis (parallel to Crawford et al. RCT)	Weekly alcohol consumption using Form 90-AQ and Steady Pattern Grid at 6 and 12 months	At 6 months significantly lower alcohol consumption in intervention vs. control group (59.7 units vs. 83.1 units, p=0.02); non-significant difference at 12 months (56.2 vs. 67.2)	Limited baseline data collected; small sample size (usable sample of 48% only); lower level of economic follow-up data than clinical outcome data. Personal costs incurred by patients not included in analyses, nor was cost of screening. Impact of variation in effectiveness measure (alcohol consumption) on ICER not examined. Service use and productivity losses evaluated by self-report – not validated against records
			Treatment cost (BI (Brief intervention) session, paperwork, onward referral)	£19; total cost per patient only £6 as only 41 of 131 (33%) in intervention group attended session
			Cost of resource utilisation (alcohol services, hospital services, primary care, other social and non-statutory services, criminal justice) and productivity losses (days off work due to alcohol misuse) all via patient self-report. Costs from local and national sources. At 6 and 12 months	At 6 months average total cost of £3068 in intervention vs. £3122 in control group (p=0.95). At 12 months £5454 versus £5207 respectively (p=0.85). Biggest proportion of costs borne by health sector (49% and 53% in intervention and control group respectively) and social services (38% and 36% in intervention and control group respectively)
			Cost-effectiveness (incremental cost-effectiveness ratio ICER) and cost-effectiveness acceptability curve	ICER = £22 per unit reduction in weekly alcohol consumption. Cost-effectiveness acceptability curve demonstrated ≥ 65% probability that brief intervention is more cost-effective than control treatment
Gentilello LM et al 2005 USA	Patients treated in ED or admitted to hospital after sustaining an injury; >=18 years; BAC (Blood alcohol concentration) >=100mg/dL or positive result on standard screening questionnaire (e.g. CAGE, MAST- Michigan Alcohol Screening Test, AUDIT) and consenting to intervention. No psychiatric illness or severe disability precluding participation in brief intervention	Cost-benefit analysis	Proportion of eligible patients (according to criteria defined in 'patient group') via literature review	27% of all injured adult patients treated in the ED	Analysis restricted to direct medical costs – i.e. from perspective of hospitals, insurers and government agencies responsible for healthcare costs. Cost of follow-up care not included. Effectiveness data based on results of a single trial
			Cost of screening and brief intervention (Medicare fees)	Screening: $16/patient (cost of BAC and forms); BI: $38/ patient (fees for BI session and administration)
			ED visits and hospitalisation rates for injury in problem-drinkers in a given year (former via literature review; latter based on national figures for ED patients requiring hospitalisation)	28% of problem-drinkers return to ED for treatment of a new injury within 1 year. 6% require hospitalisation
			Cost of ED visits and hospitalisation derived from MarketScan database of commercial claims	ED visit:$440; hospitalisation for average length of stay of 5.1 days: $16,852
			Effectiveness of BI in the context of acute injury at reducing subsequent health care use (via clinical Gentilello et al RCT)	47% reduction in new injuries requiring ED treatment or admission; 48% reduction in injuries requiring hospitalisation over 3years follow-up
			Cost analysis of universal screening and brief intervention (SBI) over 3 years	If SBI is offered, expected cost of screening, intervention and subsequent ED visits and hospitalisation = $600/patient. If BI is not offered, expected cost of subsequent ED visits and hospitalisation = $689/patient.
			Cost savings analysis	Cost saving of $89 per patient screened, or $330 per patient offered BI. BI resulted in $3.81 in health care costs saved for every $1 spent on SBI
			Sensitivity analyses (Monte Carlo analysis)	In 95% of simulations BI was either cost-saving or cost less than $24 per patient screened
Kunz FM et al. 2004 USA	295 ED patients aged >=18 years screening positive to CAGE (score >=1); spoke English or Spanish; consented; no professional alcohol counselling within past 12 months; not in police custody; medical treatment and/or cognitive impairment not precluding interview. 151 randomised to standard care (SC), 144 to brief intervention (BI) SC – standard care plus packet of health information BI – SC plus 15-20min interview by health promotion advocate. Established rapport with patient; discussed alcohol use; explored pros and cons of use (decisional balance) using reflective listening; assessed readiness to change; and negotiated plan for change based on the patient's perception of readiness. Participants given copy of plan for change Usable sample of 104 control group and 90 intervention group with follow-up data available at 3 months	Cost-effectiveness analysis (parallel to Bazargan-Hejazi et al. RCT)	Alcohol consumption using AUDIT (Alcohol Use Disorders Identification Test) score, average number of weekly drinks and heavy drinking in past month at baseline and 3 months	Intervention group had lower follow-up AUDIT score vs. control group (difference = 2.45); consumed fewer drinks per week (difference = 2.89 drinks/week); and were less likely to report heavy episodic drinking (difference = 10.34% points less). Regression analysis after controlling for baseline variables showed intervention was non-significant predictor of all 3 outcomes	Include weaknesses of RCT [i.e. selection bias as 'randomisation' not truly random (by alternative allocation) and high refusal rate (40%). Sampling period limited to 9am-6pm week days. High loss to follow-up (37%). Small sample size. Short follow-up period. Patients not blinded; investigators not blinded? Only one outcome measured]. High loss to follow-up (34%). Wide variation in ICER with changes in effectiveness value (alcohol consumption). Only cost of SBI program analysed - cost of using other services / resources not examined
			Cost of screening and brief intervention program (SBI) using DATCAP (The Drug Abuse Treatment Cost Analysis Program) - data collection instrument and interview guide; measures accounting costs (actual expenditures of a treatment program and depreciation of its resources) and economic costs (full value of all resources i.e. opportunity costs)	$631.89 per SBI participant; 79% on screening; 21% on intervention. Total program cost $91, 624; 60% on personnel salaries and benefits; 35% on overhead and patient incentives; 5% on miscellaneous supplies and equipment
			Cost-effectiveness (ICER - incremental cost-effectiveness ratio)	Cost-effectiveness ratio 257.9 for AUDIT score; 218.7 for average weekly number of drinks; 61.11 for heavy drinking

Comment(s)

The parallel cost-effectiveness analysis of a UK-based trial revealed a cost of BI of just £19 per patient. (Barrett et al, 2006). The total 12-month cost (including cost of BI, other service use and loss in productivity) amounted to £5454 for the intervention group versus £5207 for the information-only control group. The cost-effectiveness ratio was £22 per unit reduction in weekly alcohol consumption and an acceptability curve demonstrated a greater than 65% probability that referral for BI was more cost-effective than control treatment. The parallel cost-effectiveness analysis of a US-based trial demonstrated that the cost of screening and intervention (SBI) was relatively low at $631.89 per person; cost-effectiveness ratios at 3 months ranged from 61 to 258 across three measures of alcohol consumption, indicating that SBI was potentially cost-effective over the longer term. (Kunz et al, 2004) A US-based cost-benefit analysis demonstrated net cost savings of $89 per patient screened, or $330 per patient offered intervention; the benefit in reduced trauma recidivism and healthcare use produced savings of $3.81 for every $1 spent. (Gentilello et al, 2005)

Clinical Bottom Line

Cost-effectiveness and cost-savings analyses are limited by the robustness of the respective trials upon which they are based. More evidence is required, but preliminary results show that brief intervention is cost-effective in patients presenting to the Emergency Department with alcohol-related events

References

1. Barrett B, Byford S, Crawford MJ et al Cost-effectiveness of screening and referral to an alcohol health worker in alcohol misusing patients attending an accident and emergency department: A decision-making approach Drug Alcohol Depend 2006;81:47-54 2006;81:47-54
2. Gentilello LM, Ebel BE, Wickizer TM et al Alcohol interventions for trauma patients treated in emergency departments and hospitals: a cost benefit analysis Ann Surg 2005;241:541-50
3. Kunz FM, French MT, Bazargan-Hejazi S Cost-effectiveness analysis of brief intervention delivered to problem drinkers presenting at an inner-city hospital emergency department J Stud Alcohol 2004;65:363-70