ArticlePDF Available

Reducing Problematic Alcohol Use in Employees: Economic Evaluation of Guided and Unguided Web-Based Interventions Alongside a Three-arm Randomized Controlled Trial



Aims: To perform an economic evaluation of guided and unguided internet-based interventions to reduce problematic alcohol consumption in employees compared with a waitlist control condition (WLC) with unrestricted access to treatment-as-usual. Design: A cost-effectiveness analysis (CEA) and cost-utility analysis (CUA) from a societal and a cost-benefit analysis from the employer's perspective with a 6-month time horizon. Setting: Open recruitment in the German working population. Participants: Employees (178 males, 256 females, mean age 47 years) consuming at least 14 (women) or 21 (men) standard units of alcohol (SUAs) per week and scoring ≥ 8 (men) or 6 (women) on the Alcohol Use Disorders Identification Test. Measurements: Online questionnaires administered to assess SUAs, quality of life (AQoL-8D) and resource use. Outcome measure was responder (≤14/≤21 SUAs) for the CEA and quality-adjusted life years (QALYs) for the CUA. Net benefit regression was used to estimate cost-effectiveness for each study arm. Bootstrapping and sensitivity analyses were performed to account for uncertainty. Interventions: Five weekly modules including personalized normative feedback, motivational interviewing, goal setting, problem-solving and emotion regulation, provided with adherence-focused guidance (n=142; responders: n=73 [51.4%]; QALYs=0.364, SE=0.006) or without guidance (n=146; n=66 [45.2%]; 0.359, 0.007). Controls were on a waiting list (n=144; n=38 [26.4%]; 0.342, 0.007). Findings: From a societal perspective, the guided intervention had a probability of 55% (54%) of being the most efficient strategy at a willingness-to-pay (WTP) of €0 per responder (QALY) gained, compared with the unguided intervention and the control condition. At a WTP of €20,000 per QALY gained, the probability was 78%. From an employer's perspective, the guided intervention had the highest probability of a positive return on investment (81%), compared with the unguided intervention and the control condition. Conclusion: A guided internet-based intervention to reduce problematic alcohol consumption in employees appears to be both cost-beneficial and cost-effective.
Reducing problematic alcohol use in employees: economic
evaluation of guided and unguided web-based interventions
alongside a three-arm randomized controlled trial
Claudia Buntrock
| Johanna Freund
| Filip Smit
| Heleen Riper
Dirk Lehr
| Leif Boß
| Matthias Berking
| David Daniel Ebert
Department of Clinical Psychology and
Psychotherapy. Institute of Psychology,
Friedrich-Alexander University of Erlangen-
Nürnberg, Erlangen, Germany
Trimbos Institute (Netherland Institute of
Mental Health and Addiction), Utrecht, the
Department of Epidemiology and
Biostatistics, Amsterdam Public Health
Research Institute, Amsterdam University
Medical Centers, Amsterdam, the Netherlands
Department of Clinical, Neuro and
Developmental Psychology, Amsterdam Public
Health Research Institute, VU University,
Amsterdam, the Netherlands
Telepsychiatric Centre, University of
Southern Denmark, Odense, Denmark
Department of Health Psychology and
Applied Biological Psychology, Institute of
Psychology, Leuphana University Lüneburg,
Lüneburg, Germany
Department of Sport and Health Sciences,
Technical University of Munich, Munich,
Dr Claudia Buntrock, Friedrich-Alexander-
University Erlangen-Nuremberg, Department
of Clinical Psychology and Psychotherapy,
Nägelsbachstrasse 25a, 91052, Erlangen,
Funding information
European Union, Grant/Award Numbers:
2007DE161PR001, ZW6-80119999
Aims: To perform an economic evaluation of guided and unguided internet-based inter-
ventions to reduce problematic alcohol consumption in employees compared with a
waiting-list control condition (WLC) with unrestricted access to treatment-as-usual.
Design: A cost-effectiveness analysis (CEA) and costutility analysis (CUA) from a
societal and a costbenefit analysis from the employers perspective with a 6-month time
Setting: Open recruitment in the German working population.
Participants: Employees (178 males, 256 females, mean age 47 years) consuming at least
14 (women) or 21 (men) standard units of alcohol (SUAs) per week and scoring 8 (men)
or 6 (women) on the Alcohol Use Disorders Identification Test.
Measurements: On-line questionnaires administered to assess SUAs and assess quality
of life (AQoL-8D) and resource use. Outcome measure was responder (14/21 SUAs)
for the CEA and quality-adjusted life years (QALYs) for the CUA. Net benefit regression
was used to estimate cost-effectiveness for each study arm. Bootstrapping and sensitiv-
ity analyses were performed to account for uncertainty.
Interventions: Five weekly modules including personalized normative feedback, motiva-
tional interviewing, goal setting, problem-solving and emotion regulation, provided with
adherence-focused guidance [n= 142; responders: n= 73 (51.4%); QALYs = 0.364, stan-
dard error (SE) = 0.006] or without guidance [n= 146; n= 66 (45.2%); 0.359, 0.007].
Controls were on a waiting-list [n= 144; n= 38 (26.4%); 0.342, 0.007].
Findings: From a societal perspective, the guided intervention had a probability
of 55% (54%) of being the most efficient strategy at a willingness-to-pay (WTP)
of 0 per responder (QALY) gained, compared with the unguided intervention and
the control condition. At a WTP of 20 000 per QALY gained, the probability
was 78%. From an employers perspective, the guided intervention had a higher
probability of a positive return on investment (81%) compared with the unguided
intervention (58%).
Received: 9 December 2020 Accepted: 27 September 2021
DOI: 10.1111/add.15718
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any
medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2021 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction.
Addiction. 2021;112. 1
Conclusion: A guided internet-based intervention to reduce problematic alcohol
consumption in employees appears to be both cost-beneficial and cost-effective.
Cost-effectiveness, costutility, economic evaluation, employees, internet-based intervention,
problematic alcohol consumption, QALY
Alcohol has a major impact on public health. Alcohol misuse leads to a
large burden of disease, including cardiovascular diseases, mental
health conditions, digestive diseases, cancer and injuries [1]. World-
wide, approximately 3.8% of all deaths and 4.6% of disability-adjusted
life-years (DALYs) are attributable to alcohol [2]. By 2030, alcohol use
disorders (AUDs) are estimated to be the fourth leading cause of dis-
ability in high-income countries [3].
Consequently, alcohol use is associated with substantial economic
costs for society (e.g. health-care, law enforcement, social and indirect
costs stemming from productivity losses). In middle- and high-income
countries these costs account for approximately 1% of the gross
domestic product (GDP) [2]. Approximately half the socio-economic
costs (e.g. 0.64% of the GDP per country annually) are attributable to
sick leave, reduced job performance, early retirement, involuntary
unemployment and premature mortality [4].
Hence, programmes directed at employees to reduce problematic
drinking can potentially benefit the employee, the employer and soci-
ety as a whole. Problematic drinking refers to alcohol consumption
that is likely to lead to physical or psychosocial harm and is defined as
an average rate of consumption of more than 14 weekly standard
units of alcohol (SUAs, 10 g of ethanol) for women and more than
21 weekly SUAs for men [5]. Evidence suggests that screening and
brief interventions are effective at reducing excessive alcohol con-
sumption [6]. However, it seems unlikely that brief interventions alone
curb the prevalence of problem drinking [7].
Low-threshold internet- and mobile-based interventions (IMIs)
seem to be a promising option, by which evidence-based interven-
tions designed to reduce alcohol-related problems could be delivered
less intrusively [8,9]. In addition, IMIs have the potential of attracting
individuals who would otherwise not make use of traditional services
due to practical concerns or time constraints [10]. In particular, IMIs
can be anonymously accessed whenever required: two factors that
are especially relevant for problematic drinking [11].
Based on an individual participant data meta-analysis [12], IMIs
for adult problem drinking have been shown to be effective in reduc-
ing the weekly consumption of SUAs [5.02 SUAs, 95% confidence
interval (CI) = 7.57 to 2.48, P< 0.001]. Guided IMIs seemed to
yield better outcomes than unguided IMIs (6.78 SUAs, 95% CI =
12.11 to 1.45, P= 0.013) [12].
With respect to the economic merit of IMIs for problematic drink-
ing, a modelling study revealed that the implementation of IMIs could
substantially increase the cost-effectiveness of health-care systems
for AUDs [13]. A recent systematic review on the cost-effectiveness
of IMIs for substance use disorders suggested that IMIs for AUDs pro-
vide good value for money, from both a public health-care and a soci-
etal perspective [14]. The only study comparing an unguided and a
guided IMI for problematic drinking in adults in a substance abuse
treatment centre suggested that a guided IMI offered better value for
money than an unguided IMI [15].
However, to the best of our knowledge, no study has yet evalu-
ated the economic merit of (un-)guided IMIs for problematic drinking
specifically in employees, from neither the societal nor the
employers perspective. Elsewhere we have reported the primary
outcome with regard to the reduction of self-reported quantity of
alcohol consumption in standard units of alcohol [16]. Here, we
evaluated the cost-effectiveness and costutility of an unguided and
guided IMI for problematic alcohol use in employees relative to a
waiting-list control condition from a societal perspective and the
costbenefit from the employers perspective, within a time horizon
of 6 months.
Study design
The execution and reporting of the health economic evaluation
followed the declaration of the Consolidated Health Economic Evalua-
tion Reporting Standards [17] and the guidelines of the International
Society for Pharmacoeconomics and Results Research [18]. The eco-
nomic evaluation was conducted alongside a three-arm pragmatic ran-
domized controlled trial evaluating the effects of both an internet-
based intervention with adherence-focused guidance and without
guidance (i.e. self-help) to reduce alcohol consumption in employees
compared to a waiting-list control group with unrestricted access to
treatment-as-usual (TAU). Detailed information regarding the study
design can be found elsewhere [16,19]. The study was approved
by the University of Lüneburg (Germany) ethics committee
(no. Boss201404_OT) and registered in the German clinical trial regis-
ter for clinical studies (DRKS00006105).
Participants were recruited in Germany during the period from
October 2014 to February 2016. An open recruitment procedure was
used (e.g. print newspaper articles, open-access websites), which
was supported by some German health insurance companies
(e.g. BARMER, KKH, BKK) via announcements in their membership
magazines and on their websites). The recruitment strategy employed
in this study is the same that will be employed when the intervention
is delivered under real-world conditions, thus strengthening the eco-
logical validity of the study.
Individuals (aged 18 years and above) were included into the study
if they were (a) currently (self-)employed, (b) reported drinking at least
14 (women)/21 (men) standardized units of alcohol (SUAs) per week
with (c) having a score of 8/6 for men/women on the Alcohol Use
Disorders Identification Test (AUDIT) [20]. Exclusion criteria included
(a) any past psychosis or drug dependence (self-disclosed); (b) notable
suicidal risk, as indicated by a score greater than 1 on item 9 of the
Beck Depression Inventory [21]; or (c) current treatment for alcohol-
related problems or work-related stress. The Consolidated Standards of
Reporting Trials (CONSORT) study flow-chart and participants
characteristics at baseline can be found elsewhere [16]. In brief,
434 employees were recruited into the trial. Two participants withdrew
from the study and requested their data to be deleted. Thus, the final
ITT sample consisted of 432 participants, with 146 randomized to the
unguided intervention, 142 to the guided intervention and 144 to the
waiting-list control condition. At post-treatment, 339 participants
(78.5%) were still participating, whereas at 6-month follow-up, 270 par-
ticipants (62.5%) completed the follow-up questionnaires. The three
groups differed with regard to missing data on primary and secondary
outcomes at T2 (P= 0.032), but not at T3 (P= 0.092) [16]. The average
participant was female (n= 256, 59.5%; males: n= 178, 40.5%),
47 years of age [standard deviation (SD) = 19], full-time employed
(69.7%), with an average working experience of 23 years (SD = 11) and
drinking 29.6 SUAs weekly (SD = 15.8) [16].
Randomization and masking
Study participants were randomly assigned at individual level in a
1:1:1 ratio with a block size of three to the study groups by an inde-
pendent researcher, who was not otherwise involved in the study,
using an automated, computer-based, random integer generator (ran- Detailed information about the randomization proce-
dure can be found elsewhere [16]. During the randomization
procedure, group allocation was concealed from participants,
researchers involved in recruitment and eCoaches. After randomiza-
tion, study participants were aware of their group allocation as they
received immediate or delayed access to the internet-based
All study participants had unrestricted access to TAU. The German
S3-Guideline for Alcohol-related Disorders recommends brief inter-
ventions in outpatient settings for problematic drinking [e.g. general
practitioners (GPs), psychotherapists] [22]. In our pragmatic study, we
did not interfere in TAU. Instead, we maintained a naturalistic TAU
condition to represent current routine care as best as possible.
It should also be noted that health-care use was measured in detail
(see Measures).
Web-based intervention
The web-based alcohol intervention (GET.ON Clever weniger trinken;
CWT be smart drink less) consisted of five weekly modules which
were based on evidence-based treatments of alcohol use disorders
[23,24], e.g. motivational interviewing, methods to control drinking
behaviour and relapse prevention. In addition, the intervention con-
tained elements of emotion regulation [25]. A detailed description of
the intervention can be found elsewhere [19]. Participants in the
guided condition were supported by an eCoach (e.g. a trained psy-
chologist). Guidance in this study was primarily based on the
supportive-accountability model of guidance in internet interventions
[26] and consisted of two elements: adherence monitoring and feed-
back on demand, which was provided within 48 hours. Participants in
the guided intervention group completed three modules on average,
while participants in the unguided intervention group completed 2.5
modules [16].
Outcome measurements
Health-related outcomes
Drinking outcomes were operationalized as the number of partici-
pants who complied with the low-risk guideline for problematic drink-
ing at 6-month follow-up. Responders were defined as having
consumed no more than 14 (for women) or 21 (for men) SUAs weekly.
Health-related quality of life was measured using the Assessment
of Quality of Life (AQoL-8D) at baseline, post-treatment and 6-month
follow-up. The AQoL-8D is a reliable and validated quality of life
instrument [27]. It measures health-related quality of life across
eight dimensions (independent living, relationships, mental health,
coping, pain, senses, self-worth and happiness) and generates patient
preference-based utilities on a scale of 0 (death) to 1 (perfect health),
using the time trade-off method [28]. Quality-adjusted life-years
(QALYs) gained were estimated by calculating the area under the
curve (AUC) of linearly interpolated AQoL-8D utilities between mea-
surement points to cover the whole 6-month follow-up period.
Resource use and costing
Costs were measured from both societal and the employers perspec-
tives. When the societal perspective was applied, all costs
(i.e. intervention, health care, patient and family and productivity
costs) related to the intervention were taken into account irrespective
of who pays or benefits from them. When applying the employers
perspective, only costs and economic benefits pertinent to employers
were included (i.e. costs or cost reductions stemming from changes in
absenteeism and presenteeism) plus intervention costs assuming that
the latter would be paid for by the employer. We used the Trimbos
and iMTA questionnaire for costs associated with psychiatric illness
(TiC-P) [29], a retrospective questionnaire with a 3-month recall
period, for collecting data on health-care use, patient and family costs
and productivity costs. Accumulated costs were estimated using the
AUC method to linearly interpolate 3-month costs as measured at
each measurement point to cover the full follow-up period of
6 months. The TiC-P was adapted for use in Germany and has been
used in a series of cost-effectiveness studies [3032]. Costs were
expressed in Euros and indexed from 2011 to 2015, the year the
study was conducted, with an index factor of 1.05 based on the Ger-
man consumer price index [33]. Costs were converted to pounds ster-
ling (£) using the purchasing power parities reported by the
Organization for Economic Cooperation and Development [34]. For
the reference year 2015, 1 was equated to £0.89.
Intervention costs
At the time of conducting the study, the market price of the unguided
internet-based intervention provided by the GET.ON Institute, a com-
mercial health-care service provider, was 79 (£70) per participant,
whereas it was 189 (£168) for the guided intervention including the
time that eCoaches spent on coaching and administrative tasks, costs
for website maintenance and hosting, technical support and
Health-care costs
We used two German guidelines for calculating health-care costs
[35,36]. Health-care costs on a per-participant level were based on
available lists of unit cost prices [36]. Unit cost prices were as follows:
21.06 (£18.74) for a visit to the GP, 46.96 (£41.79) for a session
with a psychiatrist, 81.98 (£72.96) for a session with a psychothera-
pist and 17.14 (£15.26) for allied health services. Hospital stays were
computed at 356.70 (£317.46) for an inpatient day in a psychiatric
hospital (Supporting information, Table S2). Costs were estimated by
multiplying the units of resource use with corresponding unit cost
prices. The costs of prescribed medication were based on the German
drug registry, Rote Liste [37].
Patient and family costs
Out-of-pocket payments were directly obtained from participants.
Costs for travelling were valued at 0.30 (£0.27) per kilometre. Pro-
ductivity losses from unpaid work (e.g. household chores, shopping,
child care) were valued using the replacement cost method [38,39]
with an estimated value of 19.25 (£17.13) per hour (i.e. the average
gross wage of domestic help per hour).
Productivity costs
We followed the human capital approach to value costs due to
absenteeism [40]. Lost working days due to absenteeism were valued
at the gross average income of participants per day. Lost working
days due to presenteeism were computed by taking into account the
number of working days for which the participant reported reduced
functioning weighted by an inefficiency score for those days
(Osterhaus method) [41].
Statistical analysis
The study was not powered to statistically test differences in
health economic outcomes. Therefore, we took a probabilistic
decision-making approach to make health economic inferences
(e.g. cost-effectiveness acceptability curves) [42] and did not test
for statistically significant differences in costs between study
groups. Due to the 6-month time horizon, we did not discount
costs and effects.
Handling missing data
All findings were reported in accordance with the CONSORT [43], fol-
lowing the intention-to-treat principle. Littles overall test of random-
ness indicated that missingness in cost and outcome data occurred
completely at random (P= 0.57). We employed a Markov chain Monte
Carlo multivariate imputation algorithm as implemented in SPSS ver-
sion 26, with 100 estimations per missing value. We did not impute
costs due to inpatient care, because only six participants (1.6%) were
hospitalized during the 6-month follow-up period leading to unstable
imputations. Data were first aggregated over the 100-fold imputa-
tions and these aggregated data were used in the bootstrapped
Analyses of health-related outcomes and costs
We tested for group differences in the number of responders using
analysis [16]. Total adjusted QALYs were estimated
using ordinary least-squares regression analyses with robust stan-
dard errors controlling for AQoL-8D baseline scores [44]. Cost cat-
egories as well as costs from the employers and societal
perspectives per study group were assessed by bootstrapping
(n= 2500) ordinary least-squares regression models. In addition, we
estimated total societal costs with a generalized linear regression
model. We used the modified Parks test [45] to determine the
family distribution (i.e. gamma distribution). The model was adjusted
for baseline costs [46], initial depressive symptom severity and
alcohol consumption as associated factors of resource utilization.
We used a link identity function providing additive effects of
covariates [47].
Cost-effectiveness and costutility analyses from the
societal perspective
Net monetary benefit (NMB) regression framework was used to
obtain cost-effectiveness and costutility estimates for each condition
from the societal perspective. All three conditions were included
simultaneously in the NMB analyses, with no need to specify the com-
parator [48]. The NMB was calculated as λ×EkCk, where Ek is the
arithmetic mean of health-related outcomes (e.g. responders, QALYs),
Ck is the mean of costs for the kth comparator and λis the
willingness-to-pay (WTP) threshold. NMB values were calculated at
each WTP (QALYs: 050 000 at 5000 intervals; responder: 0
5000 at 500 intervals). At each threshold, 2500 bootstrap model iter-
ations of the linear regression models of the NMB adjusted for base-
line cost, utility values (only when QALYs were used), initial
depressive symptom severity and alcohol consumption as associated
factors for health-related outcomes and resource utilization were per-
formed. For an n-way comparison, the alternative with the highest net
benefit has the highest probability of being cost-effective [49]. Cost-
effectiveness acceptability curves (CEACs) were generated to assess
for each condition the probability of being the most cost-effective
alternative compared to the other two conditions over a range of
willingness-to-pay thresholds [50]. CEACs were based on the
bootstrapped regression models. In each of the bootstrap iterations,
the probability that each intervention was the most cost-effective
alternative was reported as the proportion of replicates, in which each
intervention had the highest NMB.
Costbenefit analyses from the employers perspective
Two metrics were applied: (1) net benefits (NB = benefits costs;
amount of money gained after costs are taken into account) and (2)
return-on-investment (ROI) [ROI = (benefits costs)/costs × 100%;
percentage of profit per Euros invested], where costs are defined as
intervention costs and benefits as the difference in productivity costs
between the intervention groups and the control condition. Both met-
rics were each estimated by bootstrapping a linear regression model
adjusted for baseline costs due to absenteeism and presenteeism, and
initial depressive symptom severity (n= 2500). The probability of
financial return was estimated by the proportion of positive estimates
(e.g. NB > 0, ROI > 0%).
Sensitivity analyses
To assess the robustness of the base case findings, six sensitivity ana-
lyses were performed. First, we repeated the main analyses but with-
out covariate adjustments. Secondly, we applied a more conservative
low-risk guideline for problematic drinking by defining responders as
having consumed no more than seven (for women) or 14 (for men)
SUAs weekly [51,52]. Thirdly, we performed analyses assuming
reduced effects in both intervention groups (i.e. approximately the
95% CI of weekly alcohol consumption and QALYs). Fourthly, we
applied winsorizing, where cost outliers (e.g. those above the 95th
percentile) are not removed, but their extreme values are replaced by
the value at the 95th percentile [53]. Fifthly, we assessed the impact
of inpatient care on the results of the main analyses by excluding
costs due to inpatient care from the analyses. Finally, we varied the
costs of the intervention by plus 50% to reflect uncertainties about
the actual market price, both in net monetary benefit regression ana-
lyses from the societal perspective and costbenefit analyses from
the employers perspective.
Health-related outcomes
At 6-month follow-up, both intervention groups yielded statistically
significant higher rates of response to the low-risk drinking threshold
(unguided: n= 66, 45.2%; guided: n= 73, 51.4%) compared to the
control group (n= 38, 26.4%; unguided: χ
= 11.16, P= 0.001;
guided: χ
= 18.85, P< 0.001). Total adjusted mean QALYs were
higher in the guided intervention group (0.364 QALYs: 95%
CI = 0.3590.369; SE = 0.006) compared to the unguided interven-
tion group (0.359 QALYs: 95% CI = 0.3540.364; SE = 0.007) and the
control condition (0.342 QALYs: 95% CI = 0.3370.357; SE = 0.007).
Adjusted incremental differences in QALYs between the interventions
groups and the control condition were statistically significant
[unguided: Δ(e) = 0.018 QALYs, 95% CI = 0.0100.025; guided: Δ(e)
= 0.022 QALYs, 95% CI = 0.0140.029].
Baseline costs were similar for the unguided intervention group [992
(£883), SD = 1477] and the control condition [917 (£816),
SD = 1580] but higher in the guided intervention group [1297
(£1154), SD = 2513]. Table 1 presents the bootstrapped (n= 2500)
imputed mean cumulative per-participant costs (in ) by condition dur-
ing the 6-month follow-up period. Direct medical and patient and
family costs were comparable for all three groups. In both intervention
groups, costs due to presenteeism were lower compared to costs cau-
sed by absenteeism. The opposite applied to the control condition.
With regard to costs stemming from absenteeism, both intervention
groups showed similar (unguided: 661, 95% CI = 462860; guided:
670, 95% CI = 467872), but higher cost levels compared to the
control condition (561, 95% CI = 360761). The guided interven-
tion group generated the fewest costs due to presenteeism (510,
95% CI = 352667) compared to the unguided intervention group
(648, 95% CI = 492803) and the control condition (628, 95%
CI = 472785). The control condition and the unguided intervention
group showed comparable cost levels [1782 (£1586), 95% CI =
14352130 versus 1774 (£1579), 95% CI = 14292119]; however,
both groups were less costly than the guided intervention group
[1954 (£1739), 95% CI = 16042303]. However, adjusted total costs
were nearly identical for both intervention groups and lower com-
pared to the control condition (Table 2).
Cost-effectiveness and costutility analyses from the
societal perspective
The control condition yielded the smallest effects in terms of treat-
ment response and QALYs gained and did so at higher costs compared
to both intervention groups, reflected in the lowest mean NMBs
(Table 2). The CEACs (Fig. 1) showed that the guided intervention
tends to be the preferred alternative compared to the unguided inter-
vention and the control condition, with a probability of 55 and 54% of
being the most cost-effective strategy at a WTP of 0 per responder
and QALY gained, respectively. The probability increases to 78%
when increasing the WTP to 20 000 (£17 800) per QALY gained and
86% at a WTP of 5000 (£4450) per additional responder. Despite
the 31 and 32% probability of the unguided intervention being the
most cost-effective strategy at a WTP of 0 per responder and QALY
gained, respectively, its probability diminishes with increasing WTPs
(e.g. 22% at a WTP of 20 000 per QALY gained). The control
condition has the lowest chance of being the most cost-effective
strategy, with a probability of 14% for both health outcomes at a
WTP of 0 that decrease to 0% as WTPs increase.
Costbenefit analyses from the employers
The unguided intervention condition showed a net benefit per partici-
pant of 29 (£26) (95% CI = 2334), which was 109 (£97) (95% CI=
105114) in the guided intervention condition. The ROI was 36%
(95% CI = 3043%) and 58% (95% CI = 5560%), respectively. The
probability of a positive return on investment was 58% for the
unguided and 81% for the guided intervention condition (Table 3).
Sensitivity analyses
Results of the sensitivity analyses are summarized in Supporting infor-
mation, Table S1. Analyses based on linear regression models without
covariate adjustments supported the conclusion that the guided inter-
vention has the highest probability of being cost-effective at a WTP
TABLE 1 Bootstrapped (n= 2500) imputed mean cumulative per-participant costs (in ) by condition over a 6-month follow-up period
Waiting-list control
condition (n= 144)
Unguided intervention
group (n= 146)
Guided intervention
group (n= 142)
Mean, (95% CI) Mean, (95% CI) Mean, (95% CI)
Intervention costs 79 189
Direct medical costs
GP 40 3447 42 3548 43 3649
Mental health care 64 3792 46 1873 70 4398
Antidepressants 15 623 19 1027 21 1229
Allied health services
36 1953 30 1346 48 3165
In-patient care 183 56310 27 0153 126 0253
Patient and family costs
Over-the-counter drugs 24 1830 22 1628 23 1729
Out-of-pocket expenses
42 2459 19 136 34 1752
Travel 8 510 7 510 4 26
Unpaid work 183 116250 175 109242 217 150285
Productivity costs
Absenteeism 561 360761 661 462860 670 467872
Presenteeism 628 472785 648 492803 510 352667
Employers perspective
Intervention costs + productivity costs 1189 8971481 1388 10981678 1368 10741662
Societal perspective
Total societal costs
1782 14352130 1774 14292119 1954 16042303
Costs of cost categories were estimated based on bootstrapped (n= 2500) linear regression models.
For example, physiotherapist, massage, occupational therapist.
For example, allied health services without prescription.
Includes all cost categories. Total societal costs were estimated based on a bootstrapped (n= 2500) linear regression model. Due to rounding, columns do
not add up correctly. CI = confidence interval.
of 20 000 per QALY gained, however; the probability was lower
(55%) compared to the adjusted analysis (78%) (Supporting
information, Fig. S1). The application of stricter limits for problematic
drinking led to converging probabilities for the guided (56%) and
unguided intervention (44%) to be the most cost-effective preventive
intervention at a WTP of 5000 (Supporting information, Fig. S2),
assuming that reduced effects on health-related outcomes in both
intervention groups did not influence the economic outcomes
(Supporting information, Fig. S3). Winsorizing extreme values to the
level at the 95th percentile did not affect cost-effectiveness outcomes
(Supporting information, Fig. S4). Hospital costs were higher in the
guided intervention group compared to the unguided group and the
control condition, so excluding these costs increased the guided inter-
ventions probability to be the most efficient option to 72% (85%) at a
WTP of 0(20 000) per QALY gained (Supporting information,
Fig. S5). Increasing intervention costs by 50% led to an almost equal
likelihood that the unguided (46%) and guided interventions (41%)
constitute the most efficient option from the societal perspective at a
WTP of 0 per QALY gained. At a WTP of 20 000 per QALY gained,
the probability of being cost-effective was higher for the guided
intervention (64%) compared to the unguided intervention (36%)
(Supporting information, Fig. S6). ROI was negative for the unguided
intervention when intervention costs were increased by 50%, while
the probability of a positive financial return was just greater than 50%
for the guided intervention group (Table 3).
Main findings
Our study was designed to evaluate the cost-effectiveness and cost
utility of the unguided and guided intervention as adjunct to TAU to
reduce problematic alcohol use in employees, in comparison with a
waiting-list control condition (WLC) with unrestricted access to TAU
from a societal and an employers perspective. Statistically significant
differences favouring both intervention groups compared to the WLC
were found for both health outcomes (e.g. treatment response and
QALYs). From a societal perspective, the guided intervention had the
highest probability of being cost-effective (e.g. 78% at a WTP of
20 000 per QALY gained). From an employers perspective, the
guided intervention showed higher net benefits than the unguided
intervention and the WLC. Probability of financial return ranged from
58% (unguided IMI) to 81% (guided IMI).
Comparison to previous research
A systematic review provided evidence that screening and brief inter-
ventions in primary care are cost-effective in relation to various com-
parators to tackle alcohol-related harms [54]. Although the
effectiveness of IMIs for adult problem drinking is well established
[12], there is a critical gap in health economic evidence for such
TABLE 2 Results from the societal perspective of adjusted cost-effectiveness and costutility analyses based on 2500 bootstrapped net monetary benefit regression models
Adjusted mean
total societal
costs ()
95% CI
responder (n) (%)
95% CI
Adjusted mean
NMB at WTP of
95% CI
Adjusted mean
NMB at WTP of
20 000/QALY gained
95% CI
Control 1942 16712213 38 26.4 0.342 0.3370.357 464 76 1004 4939 46615218
Unguided 1840 15832096 66 45.2 0.359 0.3540.364 1498 9432052 5367 50905643
Guided 1865 15922137 73 51.4 0.364 0.3590.369 1852 12902413 5498 52175778
Mean societal costs were estimated by a generalized linear regression model with gamma family distribution and link identity function adjusted for baseline costs, initial depressive symptom severity and alcohol
Quality-adjusted life-years (QALYs) were estimated by an ordinary least-squares regression model adjusted for baseline utility values.
Net monetary benefit (NMB) linear regression models were adjusted for baseline cost, initial depressive symptom severity and alcohol consumption.
Net monetary benefit (NMB) linear regression models were adjusted for baseline cost, utility values, initial depressive symptom severity and alcohol consumption. CI = confidence interval; WTP = willingness
to pay.
interventions. To our knowledge, this is the first trial-based economic
evaluation of an unguided and guided intervention to reduce problem-
atic drinking in employees using a societal and an employers perspec-
tives. As such, results from our trial add to the growing evidence
pointing to the cost-effectiveness of IMIs for mental health disorders
[14,5557]. Blankers et al. (2012) compared an unguided and a guided
IMI for harmful alcohol use in adults in a substance abuse treatment
centre from a societal perspective. Results of the current health eco-
nomic evaluation are in agreement with these findings. In parallel to
our findings, the guided IMI provided better value for money than
unguided self-help. Compared to our results, costutility analyses rev-
ealed a slightly lower probability (60%) of the guided intervention
being cost-effective compared to the unguided IMI at a ceiling ratio of
20 000 per QALY gained [15]. In addition, our findings agree with
available health economic evidence from a recent systematic review
showing the health economic benefits of IMIs for alcohol use disorder.
Probabilities that IMIs were cost-effective from a societal and a public
health care perspective, respectively, ranged from 60 to 84% [14].
Our results from the employers perspective are also in line with
findings from a recent systematic review showing that targeting sub-
stance misuse in employees improves both employeeswellbeing and
productivity [58]. Regarding the ROI analyses, our findings compare
favourably to a systematic review on the costs and benefits of health
promotion interventions at the work-place (n= 12 RCTs), which rev-
ealed on average a negative ROI (ROI = 0.22, 95% CI = 0.270.16;
min = 4.3; max = 5) [59]. In addition, the percentages of profit per
Euros invested of 37% (95% CI = 3044%) and 61% (95% CI = 58
63%) for the unguided and guided intervention, respectively,
are comparable to a study on a guided internet-based intervention
targeting work-related stress in employees (ROI = 61%) [60].
TABLE 3 Results from the employers perspective of adjusted costbenefit analyses based on 2500 bootstrapped linear regression models
(main and sensitivity analyses)
Financial return
Total 95% CI Total 95% CI NB
95% CI ROI
(%) 95% CI P
Main analyses
Unguided intervention 79 NA 108 102113 29 23 to 34 36 30 to 43 58
Guided intervention 189 NA 298 294303 109 105 to 114 58 55 to 60 81
Sensitivity analysis
Unguided intervention 118.5 NA 108 104114 10 15 to 5813 to 447
Guided intervention 283.5 NA 298 292301 14 9 to 18 5 3 to 7 54
Net benefit (NB) linear regression models were adjusted for baseline costs due to absenteeism and presenteeism and initial depressive symptom severity.
Return on investment (ROI) linear regression models were adjusted for baseline costs due to absenteeism and presenteeism, and initial depressive
symptom severity.
Probability of positive return on investment.
Intervention costs increased by 50%.
Costs are intervention costs.
Benefits are the difference in productivity costs between the intervention groups and the control condition. NA = not available; CI = confidence interval.
FIGURE 1 Cost-effectiveness acceptability
curves from the societal perspective
This study has some limitations. First, the time horizon of this study
was limited to 6 months. It is possible that health effects were
maintained after 6 months, but they also might diminish over time.
The same holds true for decreased costs and productivity gains. As
additional costs such as premature death or accidents were not taken
into account, costs only reflect short-term costs. Further studies
should thus assess the long-term clinical and cost-effectiveness of
IMIs for problematic alcohol use to shed light on its longer-term cost-
effectiveness. In addition, the societal perspective was incomplete by
omitting crime and criminal justice, future medical and opportunity
costs (e.g. time spent on using the intervention, travelling time). How-
ever, as this is a preventive intervention, crime and criminal justice
costs might not significantly affect the results in this study. Secondly,
although the sample size in this trial was sufficient to demonstrate
clinical effectiveness, it needs emphasizing that much larger sample
sizes are required for hypothesis testing in economic studies due to
the large variance of costs relative to normally distributed health
effects [61]. Therefore, future studies employing larger sample sizes
are recommended to allow for robust evaluations of cost changes and
sustainability of interventions such as IMIs. Thirdly, the IMIs were
compared to a waiting-list control condition in the present trial. How-
ever, pharmaco-economic guidelines recommend standard care
(e.g. brief face-to-face alcohol interventions) as comparator [62].
Future studies should thus directly compare the cost-effectiveness of
IMIs versus face-to-face interventions. Fourthly, the trial participants
were highly educated. Evidence suggests that better treatment adher-
ence is predicted by higher education [63]. Attrition has been
suggested to be an issue, especially in internet-delivered interventions
[64]. Hence, we cannot predict the uptake of such an intervention in
less educated people or among people with a lower socio-economic
status. It is thus warranted to conduct economic evaluations in these
specific population segments. Fifthly, we did not conduct diagnostic
interviews to identify participants with alcohol use disorder. However,
including participants with a wide range of consumed alcohol units
reflect the real-life situation in the general population in high-income
countries [65]. Finally, the research context may have led to self-
selection of individuals who might be more motivated and committed
to engage in IMIs than is assumed outside a research context [66]. As
a result, findings might not be generalizable to the wider target popu-
lation, but are likely to be representative for precisely those people
willing to use IMIs in the first place.
The current study shows that an internet-based intervention may
not only be effective in reducing weekly alcohol consumption, but
also that achieving and maintaining a marked reduction in drinking is
associated with significant increases in health-related quality of life.
As the population segment targeted in the current study had a
lower than average health-related quality of life when entering the
study [67], this finding underscores the importance of offering this
target group an eHealth intervention to curb their problematic
alcohol use.
Internet-based interventions for mental disorders have often
been touted as potential cost-saving alternatives to face-to-face
individual or group therapy [55,56]. Findings from our study add to
the evidence base that IMIs have indeed a high probability of being
cost-effective in reducing problematic alcohol consumption among
employees. The IMIs that we evaluated are cost-effective and even
dominant, in the sense that for fewer costs better health gains were
achieved. Furthermore, the outcomes of our ROI analyses could
encourage employers and decision-makers in public health to offer
IMIs to employees because there are favourable ROIs, as the IMIs
led to increases in productivity (less absenteeism and less
presenteeism), in particular via an IMI with adherence-focused guid-
ance. All in all, the findings highlight the importance of promoting
awareness and access to this type of intervention for problematic
Considering that only relatively few health-care professionals
actually administer face-to-face brief alcohol interventions, and that
only a small proportion of patients who might benefit accept those
treatment offers [7], it would be worthwhile to integrate IMIs for
problematic drinking into routine practice. However, some risks need
to be considered when scaling-up IMIs. There are no guarantees that
adherence and (by proxy) effectiveness found in a research context
will be maintained if such an intervention is scaled-up in the general
population, at the work-place or in primary care. In addition, the high-
quality information and communication technology resources
(e.g. stable and secure internet connections) may not be available to
the same extent outside the research setting.
Findings suggest that internet-based interventions to reduce problem-
atic alcohol consumption in employees are both cost-beneficial
(i.e. the financial benefits exceed the intervention costs and thus the
return on investment is positive) and cost-effective (i.e. the health
effects gained present good value for the money invested). However,
more studies with longer follow-up periods and standard care as com-
parator are needed to further substantiate these findings. Given the
evidence for the effectiveness, feasibility and acceptance of internet-
based interventions to reduce problematic alcohol consumption, their
potential cost-effectiveness and scalability might strategically pave
the way to alleviate the health and economic burden related to prob-
lematic alcohol use in an affordable manner.
C.B., J.F., F.S., H.R. and L.B. have no competing interests to disclose.
D.D.E., D.L. and M.B. are stakeholders of the GET.ON Institute, which
aims to implement scientific findings related to digital health interven-
tions into routine care. D.D.E. has served as a consultant to/on the
scientific advisory boards of Sanofi, Novartis, Minddistrict, Lantern,
Schoen Kliniken, Ideamed and German health insurance companies
(BARMER, Techniker Krankenkasse) and a number of federal
chambers for psychotherapy.
This was not an industry-supported study. The European Union
funded this study (EU EFRE: ZW6-80119999, CCI 2007DE161
PR001). The funder did not have a role in study design, data
collection, analysis and interpretation of results, or the decision to
publish the study results.
Claudia Buntrock: Conceptualization-Lead, Formal analysis-Lead,
Methodology-Lead, Writing-original draft-Lead. Johanna Freund:
Writing-original draft-Supporting, Writing-review & editing-
Supporting. Filip Smit: Conceptualization-Supporting, Formal analysis-
Supporting, Methodology-Supporting, Supervision-Supporting,
Writing-review & editing-Supporting. Heleen Riper: Conceptualiza-
tion-Supporting, Supervision-Supporting, Writing-review & editing-
Supporting. Dirk Lehr: Conceptualization-Supporting, Funding
acquisition-Equal, Project administration-Supporting, Supervision-
Supporting, Writing-review & editing-Supporting. Leif Boß: Conceptu-
alization-Supporting, Data curation-Lead, Investigation-Lead, Project
administration-Lead, Writing-review & editing-Supporting. Matthias
Berking: Conceptualization-Supporting, Funding acquisition-
Equal, Writing-review & editing-Supporting. David Daniel
Ebert: Conceptualization-Supporting, Funding acquisition-Equal,
Supervision-Supporting, Writing-review & editing-Supporting.
German Clinical Trial Registration DRKS00006105, date of registra-
tion: 2014-07-07.
Claudia Buntrock
Johanna Freund
Filip Smit
Heleen Riper
Dirk Lehr
Leif Boß
Matthias Berking
David Daniel Ebert
1. Hammer JH, Parent MC, Spiker DA, World Health Organization.
Global status report on alcohol and health 2018 [internet]. 65.
Available at:
global_alcohol_report/msbgsruprofiles.pdf. Accessed 15 Nov 2020.
2. Rehm J, Mathers C, Popova S, Thavorncharoensap M,
Teerawattananon Y, Patra J. Global burden of disease and injury and
economic cost attributable to alcohol use and alcohol-use disorders.
Lancet 2009;373:22232233.
3. Mathers C, Loncar D. Projections of global mortality and burden of
disease from 2002 to 2030. PLOS Med. 2006;3:201130.
4. Laramée P, Kusel J, Leonard S, Aubin H-J, François C, Daeppen J-B.
The economic burden of alcohol dependence in Europe. Alcohol
Alcohol 2013;48:259269.
5. World Health Organization (WHO). WHO Guide to Mental Health in
Primary Care. London, UK: Royal Society of Medicine Press; 2000.
6. ODonnell A, Anderson P, Newbury-Birch D, Schulte B, Schmidt C,
Reimer J, et al. The impact of brief alcohol interventions in primary
healthcare: a systematic review of reviews. Alcohol Alcohol. 2014;
7. McCambridge J, Saitz R. Rethinking brief interventions for alcohol in
general practice. BMJ. 2017;356:j116.
8. Ames GM, Bennett JB. Prevention interventions of alcohol problems
in the workplace: a review and guiding framework. Alcohol Res
Health. 2011;34:17587.
9. Lehr D, Geraedts A, Asplund RP, Khadjesari Z, Heber E, de Bloom J,
et al. Occupational e-mental health: current approaches and promis-
ing perspectives for promoting mental health in workers. In:
Wiencke M, Fischer S, Cacace M, editorsHealthy at WorkInterdis-
ciplinary Perspectives. Cham: Springer International Publishing;
2016. p. 25781.
10. Carlbring P, Andersson G, Cuijpers P, Riper H, Hedman-Lagerlöf E.
Internet-based vs. face-to-face cognitive behavior therapy for psy-
chiatric and somatic disorders: An updated systematic review and
meta-analysis. Cogn Behav Ther. 2018;47:118.
11. Tait RJ, Spijkerman R, Riper H. Internet and computer based inter-
ventions for cannabis use: a meta-analysis. Drug Alcohol Depend.
12. Riper H, Hoogendoorn A, Cuijpers P, Karyotaki E, Boumparis N,
Mira A, et al. Effectiveness and treatment moderators of internet
interventions for adult problem drinking: an individual patient data
meta-analysis of 19 randomised controlled trials. PLOS Med 2018;
13. Smit F, Lokkerbol J, Riper H, Majo MC, Boon B, Blankers M. Model-
ing the cost-effectiveness of health care systems for alcohol use
disorders: how implementation of eHealth interventions improves
cost-effectiveness. J Med Internet Res. 2011;13:e56.
14. Buntrock C, Kählke F, Smit F, Ebert DD. A systematic review of
trial-based economic evaluations of internet-and mobile-based
interventions for substance use disorders. Eur J Pub Health. 2021;
15. Blankers M, Nabitz U, Smit F, Koeter MWJ, Schippers GM. Economic
evaluation of internet-based interventions for harmful alcohol use
alongside a pragmatic randomized controlled trial. J Med Internet
Res. 2012;14:e134.
16. Boß L, Lehr D, Schaub MP, Paz Castro R, Riper H, Berking M, et al.
Efficacy of a web-based intervention with and without guidance for
employees with risky drinking: results of a three-arm randomized
controlled trial. Addiction. 2018;113:63546.
17. Husereau D, Drummond M, Petrou S, Carswell C,
Moher D, Greenberg D, et al. Consolidated health economic
evaluation reporting standards (CHEERS)explanation and elabora-
tion: a report of the ISPOR health economic evaluation publication
guidelines good reporting practices task force. Value Health. 2013;
18. Ramsey SD, Willke RJ, Glick H, Reed SD, Augustovski F, Jonsson B
et al. Cost-effectiveness analysis alongside clinical trials IIan ISPOR
good research practices task force report. Value Health. 2015;18:
19. Boß L, Lehr D, Berking M, Riper H, Schaub MP, Ebert DD. Evaluating
the (cost-)effectiveness of guided and unguided internet-based
self-help for problematic alcohol use in employeesa three arm
randomized controlled trial. BMC Public Health. 2015;15:1043.
20. Babor TF, Higgins-Biddle JC, Saunders JB, Monteiro MG.
The Alcohol Use Disorders Identification Test: Guidelines for Use in
Primary Care. 2nd ed. Geneva: World Health Organization; 2001.
21. Beck AT, Steer RA, Brown GK. BDI-II: Beck Depression
Inventory Manual. 2nd ed. San Antonio: Psychological Corporation;
22. Mann K, Batra A, Hoch E, die Leitliniengruppe. S3-Leitlinie
Screening, Diagnose und Behandlung alkoholbezogener Störungen.
Sucht. 2017;63:724.
23. Miller WR, Muñoz RF. Controlling Your Drinking: Tools to Make
Moderation Work for You. 2nd ed. New York: Guilford Press; 2013.
24. Hester RK, Miller WR. Handbook of Alcoholism Treatment
Approaches: Effective Alternatives. 3rd ed. Boston: Allyn and Bacon;
25. Berking M, Ebert D, Cuijpers P, Hofmann SG. Emotion regulation
skills training enhances the efficacy of inpatient cognitive behavioral
therapy for major depressive disorder: a randomized controlled trial.
Psychother Psychosom. 2013;82:23445.
26. Mohr DC, Cuijpers P, Lehman K. Supportive accountability: a model
for providing human support to enhance adherence to eHealth inter-
ventions. J Med Internet Res. 2011;13:e30.
27. Richardson J, Iezzi A, Khan MA, Maxwell A. Validity and
reliability of the assessment of quality of life (AQoL)-8D
multi-attribute utility instrument. PatientPatient-Centered OutcRes.
28. Richardson J, Sinha K, Iezzi A, Khan MA. Modelling utility weights for
the assessment of quality of life (AQoL)-8D. Qual Life Res. 2014;23:
29. Bouwmans C, De Jong K, Timman R, Zijlstra-Vlasveld M, Van der
Feltz-Cornelis C, Tan SS, et al. Feasibility, reliability and validity of a
questionnaire on healthcare consumption and productivity loss in
patients with a psychiatric disorder (TiC-P). BMC Health Serv Res.
30. Buntrock C, Berking M, Smit F, Lehr D, Nobis S, Riper H, et al.
Preventing depression in adults with subthreshold depression:
health-economic evaluation alongside a pragmatic randomized con-
trolled trial of a web-based intervention. J Med Internet Res. 2017;
31. Nobis S, Lehr D, Ebert DD, Berking M, Heber E, Baumeister H, et al.
Efficacy and cost-effectiveness of a web-based intervention with
mobile phone support to treat depressive symptoms in adults with
diabetes mellitus type 1 and type 2: design of a randomised con-
trolled trial. BMC Psychiatry. 2013;13:110.
32. Thiart H, Ebert DD, Lehr D, Nobis S, Buntrock C, Berking M, et al.
Internet-based cognitive behavioral therapy for insomnia: a health
economic evaluation. Sleep. 2016;39:176978.
33. DESTATIS. German Consumer Price Index. Available at:
AC4640517.live712. Accessed 15 Oct 2020.
34. Organization for Economic Cooperation and Development (OECD).
Exchange rates (indicator). 2015.
35. Krauth C, Hessel F, Hansmeier T, Wasem J, Seitz R, Schweikert B.
Empirical standard costs for health economic evaluation in
Germanya proposal by the working group methods in health eco-
nomic evaluation. Gesundheitswesen. 2005;67:73646.
36. Bock J-O, Brettschneider C, Seidl H, Bowles D, Holle R, Greiner W
et al. Calculation of standardised unit costs from a societal perspec-
tive for health economic evaluation. Gesundheitswesen. 2015;77:
37. Rote Liste Service GmbH. Rote Liste 2013. Frankfurt/Main: Rote
Liste Service GmbH; 2013.
38. Krol M, Brouwer W. Unpaid work in health economic evaluations.
Soc Sci Med. 2015;144:12737.
39. Koopmanschap MA, van Exel NJA, van den Berg B, Brouwer WBF.
An overview of methods and applications to value informal care in
economic evaluations of healthcare. Pharmacoeconomics. 2008;26:
40. Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW.
Methods for the Economic Evaluation of Health Care Programmes.
3rd ed. Oxford, UK: Oxford University Press; 2015.
41. Osterhaus JT, Gutterman DL, Plachetka JR. Healthcare resource
and lost labour costs of migraine headache in the US.
PharmacoEconomics. 1992;2:6776.
42. van Hout BA, Al MJ, Gordon GS, Rutten FF. Costs, effects and
C/E-ratios alongside a clinical trial. Health Econ. 1994;3:30919.
43. Eysenbach G. CONSORT-EHEALTH: improving and standardizing
evaluation reports of web-based and mobile health interventions.
J Med Internet Res. 2011;13:e126.
44. Pullenayegum EM, Tarride JE, Xie F, Goeree R, Gerstein HC,
OReilly D. Analysis of health utility data when some subjects attain
the upper bound of 1: are tobit and CLAD models appropriate? Value
Health. 2010;13:48794.
45. Manning WG, Mullahy J. Estimating log models: to transform or not
to transform? J Health Econ. 2001;20:46194.
46. van Asselt ADI, van Mastrigt GAPG, Dirksen CD, Arntz A,
Severens JL, Kessels AGH. How to deal with cost differences at
baseline. PharmacoEconomics. 2009;27:51928.
47. Barber J, Thompson S. Multiple regression of cost data: use of gener-
alised linear models. J Health Serv Res Policy. 2004;9:197204.
48. Briggs A, Sculpher M, Claxton K. Decision Modelling for Health Eco-
nomic Evaluation. Oxford, UK: Oxford University Press; 2006.
49. Edmond S-W, Grieve R, Carpenter JR. Two-stage nonparametric
bootstrap sampling with shrinkage correction for clustered data.
Stata J. 2013;13:14164.
50. Briggs AH, OBrien BJ, Blackhouse G. Thinking outside the box:
recent advances in the analysis and presentation of uncertainty in
cost-effectiveness studies. Annu Rev Public Health. 2002;23:
51. Seitz H, Bühringer G. Empfehlungen des wissenschaftlichen
Kuratoriums der DHS zu Grenzwerten für den Konsum alkoholischer
Getränke [Recommendations of the Scientific Board of Trustees of
the DHS on Limits for the Consumption of Alcoholic Beverages].
Hamm, Germany: DHS; 2010.
52. Lange C, Manz K, Rommel A, Schienkiewitz A, Mensink G. Alcohol
consumption of adults in Germany: harmful drinking quantities, con-
sequences and measures. J Health Monit. 2016;220.
53. Dixon WJ, Yuen KK. Trimming and winsorization: a review. Stat
Hefte. 1974;15:15770.
54. Angus C, Latimer N, Preston L, Li J, Purshouse R. What are the impli-
cations for policy makers? A systematic review of the cost-
effectiveness of screening and brief interventions for alcohol misuse
in primary care. Front Psychiatry. 2014;5:11424.
55. Donker T, Blankers M, Hedman E, Ljotsson B, Petrie K,
Christensen H. Economic evaluations of internet interventions
for mental health: a systematic review. Psychol Med. 2015;45:
56. Paganini S, Teigelkötter W, Buntrock C, Baumeister H. Economic
evaluations of internet- and mobile-based interventions for the treat-
ment and prevention of depression: a systematic review. J Affect
Disord. 2018;225:73355.
57. Ophuis RH, Lokkerbol J, Heemskerk SCM, van Balkom AJLM,
Hiligsmann M, Evers SMAA. Cost-effectiveness of interventions for
treating anxiety disorders: a systematic review. J Affect Disord.
58. de Oliveira C, Cho E, Kavelaars R, Jamieson M, Bao B, Rehm J. Eco-
nomic analyses of mental health and substance use interventions in
the workplace: a systematic literature review and narrative synthesis.
Lancet Psychiatry. 2020;7:893910.
59. Baxter S, Sanderson K, Venn AJ, Blizzard CL, Palmer AJ. The relation-
ship between return on investment and quality of study methodol-
ogy in workplace health promotion programs. Am J Health Promot.
60. Ebert DD, Kählke F, Buntrock C, Berking M, Smit F, Heber E, et al. A
health economic outcome evaluation of an internet-based mobile-
supported stress management intervention for employees. Scand J
Work Environ Health. 2018;44:17182.
61. Briggs A. Economic evaluation and clinical trials: size matters. BMJ.
62. Sacristán JA, Abellán-Perpiñán J-M, Dilla T, Soto J, Oliva J. Some
reflections on the use of inappropriate comparators in CEA. Cost Eff
Resour Alloc. 2020;18:2935.
63. Batterham PJ, Neil AL, Bennett K, Griffiths KM, Christensen H. Pre-
dictors of adherence among community users of a cognitive behavior
therapy website. Patient Prefer Adher. 2008;2:97105.
64. Price M, Gros DF, McCauley JL, Gros KS, Ruggiero KJ.
Nonuse and dropout attrition for a web-based mental health
intervention delivered in a post-disaster context. Psychiatry. 2012;
65. Wood AM, Kaptoge S, Butterworth A, Nietert PJ, Warnakula S,
Bolton T, et al. Risk thresholds for alcohol consumption: combined
analysis of individual-participant data for 599 912 current drinkers in
83 prospective studies. Lancet. 2018;391:151323.
66. Baumel A, Edan S, Kane JM. Is there a trial bias impacting
user engagement with unguided e-mental health interventions?
A systematic comparison of published reports and real-world usage
of the same programs. Transl Behav Med. 2019;9:102033.
67. Grochtdreis T, Dams J, König H-H, Konnopka A. Health-related qual-
ity of life measured with the EQ-5D-5L: estimation of normative
index values based on a representative German population sample
and value set. Eur J Health Econ HEPAC Health Econ Prev Care.
Additional supporting information may be found in the online version
of the article at the publishers website.
How to cite this article: Buntrock C, Freund J, Smit F, Riper H,
Lehr D, Boß L, et al. Reducing problematic alcohol use in
employees: economic evaluation of guided and unguided web-
based interventions alongside a three-arm randomized
controlled trial. Addiction. 2021;112.
... Effective internet-based interventions may reduce this gap due to their low-threshold accessibility, high scalability and reduced stigmatisation (Riper et al., 2018). Moreover, internet-and mobile-based interventions are much more likely to be more cost effective than treatment as usual (Buntrock et al., 2019;Buntrock et al., 2021). ...
Full-text available
Aims: Online interventions reduce the treatment gap between the number of people with alcohol misuse and people who actually receive help. This study investigated the effectiveness and predictors of success of a Belgian online help programme. Methods: A real-life retrospective open cohort study evaluating the guided and unguided internet intervention on the Belgian online platform The intervention consisted of a 12-week programme based upon cognitive behaviour therapy, motivational interviewing and acceptance and commitment therapy. Inclusion criteria are age above 18 years, recording of alcohol consumption in the daily journal for at least 2 weeks, and minimum 2 chat sessions in the guided group.Outcomes were weekly alcohol consumption after 6 and 12 weeks and treatment response (drinking less than 10 or 20 standard units (SU) per week). Additional analysis was done on predictors of success. Results: A total of 460 participants in the guided group and 968 in the self-help group met the inclusion criteria. Average baseline alcohol consumption in the two groups was 40 SU per week. Alcohol consumption decreased by 31 SU (Cohen's d 1.17, p
Full-text available
Abstract Although the choice of the comparator is one of the aspects with a highest effect on the results of cost-effectiveness analyses, it is one of the less debated issues in international methodological guidelines. The inclusion of an inappropriate comparator may introduce biases on the outcomes and the recommendations of an economic analysis. Although the rules for cost-effectiveness analyses of sets of mutually exclusive alternatives have been widely described in the literature, in practice, they are hardly ever applied. In addition, there are many cases where the efficiency of the standard of care has never been assessed; or where the standard of care has demonstrated to be cost-effective with respect to a non-efficient option. In all these cases the comparator may lie outside the efficiency frontier, so the result of the CEA may be biased. Through some hypothetical examples, the paper shows how the complementary use of an independent reference may help to identify potential inappropriate comparators and inefficient use of resources.
Full-text available
Background: Substance use disorders (SUDs) contribute significantly to global rates of morbidity and mortality. Internet- and mobile-based interventions (IMIs) have been suggested as an adjunct to face-to-face health services. However, the evidence for the cost-effectiveness of IMIs for SUDs is scant. Methods: A comprehensive literature search in PubMed, PsycINFO, the Cochrane Central Register of Controlled Trials, NHS Economic Evaluations Database, NHS Health Technology Assessment Database, Office of Health Economics Evaluations Database and EconLit was conducted. We included economic evaluations alongside randomized controlled trials of IMIs for SUDs compared with a control group. Results: Of 1687 abstracts identified, 11 studies met the inclusion criteria. Targeted conditions were alcohol use disorder (four studies) and tobacco smoking (five studies) whereas two studies included any SUD. Cost-effectiveness results demonstrated that IMIs had a firm probability of being more cost-effective than TAU (e.g. less costs per additional abstinent person). Compared with (online) psycho-education, evidence towards an additional benefit of IMIs was less clear. Regarding cost-utility (e.g. costs per quality-adjusted life year gained), except for one study, results suggested that TAU and online psycho-education would probably be more preferable than IMIs. Quality of study reporting was at least adequate. Conclusions: The likelihood of IMIs being more cost-effective than TAU looks promising but more economic evaluations are needed in order to determine the economic merit of IMIs. With an increasing pressure on health care budgets, strategies to disseminate effective interventions at affordable costs are required. This review suggests that IMIs might carry that promise and have potential as a cost-effective strategy to scale-up existing evidence-based treatments for SUDs. Systematic review registration: The systematic review has been registered in the PROSPERO database (no. CRD42018099486).
Full-text available
Background The generic and preference-based instrument EQ-5D is available in a five-response levels version (EQ-5D-5L). A value set for the EQ-5D-5L based on a representative sample of the German population has recently been developed. The aim of this study was to estimate normative values of the EQ-5D-5L index for Germany, and to examine associations between the EQ-5D-5L and selected sociodemographic factors. Methods The analysis was based on a representative sample (n = 4998) of the German general adult population in 2014. Participants had to rate their health-related quality of life on the EQ-5D-5L descriptive system as well as on a visual analogue scale (EQ-VAS). Normative values of the EQ-5D-5L index were estimated for selected sociodemographic characteristics. For the examination of associations between EQ-5D-5L index scores and selected sociodemographic factors, multivariate regression analyses were used. Results The mean EQ-5D-5L index score of the total sample was 0.88 (SD 0.18), corresponding to an overall mean EQ-VAS score of 71.59 (SD 21.36). Female gender and increasing age were associated with a lower EQ-5D-5L index score (p < 0.001). Higher education, full-time employment and private health insurance were associated with a higher EQ-5D-5L index score (p < 0.001). Conclusion This was the first study to estimate normative values of the EQ-5D-5L index for Germany based on a representative sample. The German normative values of the EQ-5D-5L are comparable to those reported for other countries. However, the mean EQ-5D-5L index score of the total sample was worse than those of the samples of studies from other countries.
Full-text available
Background Low-risk limits recommended for alcohol consumption vary substantially across different national guidelines. To define thresholds associated with lowest risk for all-cause mortality and cardiovascular disease, we studied individual-participant data from 599 912 current drinkers without previous cardiovascular disease. Methods We did a combined analysis of individual-participant data from three large-scale data sources in 19 high-income countries (the Emerging Risk Factors Collaboration, EPIC-CVD, and the UK Biobank). We characterised dose–response associations and calculated hazard ratios (HRs) per 100 g per week of alcohol (12·5 units per week) across 83 prospective studies, adjusting at least for study or centre, age, sex, smoking, and diabetes. To be eligible for the analysis, participants had to have information recorded about their alcohol consumption amount and status (ie, non-drinker vs current drinker), plus age, sex, history of diabetes and smoking status, at least 1 year of follow-up after baseline, and no baseline history of cardiovascular disease. The main analyses focused on current drinkers, whose baseline alcohol consumption was categorised into eight predefined groups according to the amount in grams consumed per week. We assessed alcohol consumption in relation to all-cause mortality, total cardiovascular disease, and several cardiovascular disease subtypes. We corrected HRs for estimated long-term variability in alcohol consumption using 152 640 serial alcohol assessments obtained some years apart (median interval 5·6 years [5th–95th percentile 1·04–13·5]) from 71 011 participants from 37 studies. Findings In the 599 912 current drinkers included in the analysis, we recorded 40 310 deaths and 39 018 incident cardiovascular disease events during 5·4 million person-years of follow-up. For all-cause mortality, we recorded a positive and curvilinear association with the level of alcohol consumption, with the minimum mortality risk around or below 100 g per week. Alcohol consumption was roughly linearly associated with a higher risk of stroke (HR per 100 g per week higher consumption 1·14, 95% CI, 1·10–1·17), coronary disease excluding myocardial infarction (1·06, 1·00–1·11), heart failure (1·09, 1·03–1·15), fatal hypertensive disease (1·24, 1·15–1·33); and fatal aortic aneurysm (1·15, 1·03–1·28). By contrast, increased alcohol consumption was log-linearly associated with a lower risk of myocardial infarction (HR 0·94, 0·91–0·97). In comparison to those who reported drinking >0–≤100 g per week, those who reported drinking >100–≤200 g per week, >200–≤350 g per week, or >350 g per week had lower life expectancy at age 40 years of approximately 6 months, 1–2 years, or 4–5 years, respectively. Interpretation In current drinkers of alcohol in high-income countries, the threshold for lowest risk of all-cause mortality was about 100 g/week. For cardiovascular disease subtypes other than myocardial infarction, there were no clear risk thresholds below which lower alcohol consumption stopped being associated with lower disease risk. These data support limits for alcohol consumption that are lower than those recommended in most current guidelines. Funding UK Medical Research Council, British Heart Foundation, National Institute for Health Research, European Union Framework 7, and European Research Council.
Full-text available
During the last two decades, Internet-delivered cognitive behavior therapy (ICBT) has been tested in hundreds of randomized controlled trials, often with promising results. However, the control groups were often waitlisted, care-as-usual or attention control. Hence, little is known about the relative efficacy of ICBT as compared to face-to-face cognitive behavior therapy (CBT). In the present systematic review and meta-analysis, which included 1418 participants, guided ICBT for psychiatric and somatic conditions were directly compared to face-to-face CBT within the same trial. Out of the 2078 articles screened, a total of 20 studies met all inclusion criteria. Results showed a pooled effect size at post-treatment of Hedges g = .05 (95% CI, -.09 to .20), indicating that ICBT and face-to-face treatment produced equivalent overall effects. Study quality did not affect outcomes. While the overall results indicate equivalence, there have been few studies of the individual psychiatric and somatic conditions so far, and for the majority, guided ICBT has not been compared against face-to-face treatment. Thus, more research, preferably with larger sample sizes, is needed to establish the general equivalence of the two treatment formats.
Full-text available
Objective This study aimed to estimate and evaluate the cost-effectiveness and cost-benefit of a guided internet- and mobile-supported occupational stress-management intervention (iSMI) for employees from the employer's perspective alongside a randomized controlled trial. Methods A sample of 264 employees with elevated symptoms of perceived stress (Perceived Stress Scale, PSS-10 ≥22) was randomly assigned either to the iSMI or a waitlist control (WLC) group with unrestricted access to treatment as usual. The iSMI consisted of seven sessions of problem-solving and emotion-regulation techniques and one booster session. Self-report data on symptoms of perceived stress and economic data were assessed at baseline, and at six months following randomization. A cost-benefit analysis (CBA) and a cost-effectiveness analysis (CEA) with symptom-free status as the main outcome from the employer's perspective was carried out. Statistical uncertainty was estimated using bootstrapping (N=5000). Results The CBA yielded a net-benefit of EUR181 [95% confidence interval (CI) -6043-1042] per participant within the first six months following randomization. CEA showed that at a willingness-to-pay ceiling of EUR0, EUR1000, EUR2000 for one additional symptom free employee yielded a 67%, 90%, and 98% probability, respectively, of the intervention being cost-effective compared to the WLC. Conclusion The iSMI was cost-effective when compared to WLC and even lead to cost savings within the first six months after randomization. Offering stress-management interventions can present good value for money in occupational healthcare.
Mental illness and substance use disorders in the workplace have been increasingly recognised as a problem in most countries; however, evidence is scarce on which solutions provide the highest return on investment. We searched academic and grey literature databases and additional sources for studies that included a workplace intervention for mental health or substance abuse, or both, and that did an economic analysis. We analysed the papers we found to identify the highest yielding and most cost-effective interventions by disorder. On the basis of 56 studies, we found moderate strength of evidence that cognitive behavioural therapy is cost-saving (and in some cases cost-effective) to address depression. We observed strong evidence that regular and active involvement of occupational health professionals is cost-saving and cost-effective in reducing sick leave related to mental health and in encouraging return to work. We identified moderate evidence that coverage for pharmacotherapy and brief counselling for smoking cessation are both cost-saving and cost-effective. Addressing mental health and substance misuse in the workplace improves workers' wellbeing and productivity, and benefits employers' bottom line (ie, profit). Future economic analyses would benefit from the consideration of subgroup analyses, examination of longer follow-ups, inclusion of statistical and sensitivity analyses and discussion around uncertainty, and consideration of potential for bias.
Trial settings that include proactive recruitment, human contact, and assessment procedures may substantially impact the way users engage with unguided e-mental health programs and the generalizability of reported findings. This study examined the impact of trial setting on user behavior by directly comparing reported user engagement in trial-based research and objective measures of real-world usage of the same unguided mental health programs. The authors conducted a systematic search for papers reporting user engagement with off-the-shelf unguided e-mental health programs. Real-world usage was obtained from a panel that presents aggregated nonpersonal information on user engagement with digital programs across the world. A total of 13 papers yielding 14 comparable usage metrics met all inclusion criteria. In three papers reporting the use of programs by lay users without any proactive trial procedures, the ratios calculated by dividing the usage reported in the paper by the usage documented within the objective dataset were 0.84, 1.05, and 1.27—suggesting a sufficient criterion validity for our examination. In studies that proactively recruited users and included pre- to post-assessment procedures (11 comparisons), the median program usage rate reported was 4.06 times higher (IQR = 4.49) than the real-world usage of the same program. Severity of clinical symptoms, in-person versus remote assessment procedures, study design, and program cost had no impact on these differences. The results suggest that trial settings have a large impact on user engagement with unguided interventions and, therefore, on the generalizability of the findings to the real world.
Background Face-to-face brief interventions for problem drinking are effective, but they have found limited implementation in routine care and the community. Internet-based interventions could overcome this treatment gap. We investigated effectiveness and moderators of treatment outcomes in internet-based interventions for adult problem drinking (iAIs). Methods and findings Systematic searches were performed in medical and psychological databases to 31 December 2016. A one-stage individual patient data meta-analysis (IPDMA) was conducted with a linear mixed model complete-case approach, using baseline and first follow-up data. The primary outcome measure was mean weekly alcohol consumption in standard units (SUs, 10 grams of ethanol). Secondary outcome was treatment response (TR), defined as less than 14/21 SUs for women/men weekly. Putative participant, intervention, and study moderators were included. Robustness was verified in three sensitivity analyses: a two-stage IPDMA, a one-stage IPDMA using multiple imputation, and a missing-not-at-random (MNAR) analysis. We obtained baseline data for 14,198 adult participants (19 randomised controlled trials [RCTs], mean age 40.7 [SD = 13.2], 47.6% women). Their baseline mean weekly alcohol consumption was 38.1 SUs (SD = 26.9). Most were regular problem drinkers (80.1%, SUs 44.7, SD = 26.4) and 19.9% (SUs 11.9, SD = 4.1) were binge-only drinkers. About one third were heavy drinkers, meaning that women/men consumed, respectively, more than 35/50 SUs of alcohol at baseline (34.2%, SUs 65.9, SD = 27.1). Post-intervention data were available for 8,095 participants. Compared with controls, iAI participants showed a greater mean weekly decrease at follow-up of 5.02 SUs (95% CI −7.57 to −2.48, p < 0.001) and a higher rate of TR (odds ratio [OR] 2.20, 95% CI 1.63–2.95, p < 0.001, number needed to treat [NNT] = 4.15, 95% CI 3.06–6.62). Persons above age 55 showed higher TR than their younger counterparts (OR = 1.66, 95% CI 1.21–2.27, p = 0.002). Drinking profiles were not significantly associated with treatment outcomes. Human-supported interventions were superior to fully automated ones on both outcome measures (comparative reduction: −6.78 SUs, 95% CI −12.11 to −1.45, p = 0.013; TR: OR = 2.23, 95% CI 1.22–4.08, p = 0.009). Participants treated in iAIs based on personalised normative feedback (PNF) alone were significantly less likely to sustain low-risk drinking at follow-up than those in iAIs based on integrated therapeutic principles (OR = 0.52, 95% CI 0.29–0.93, p = 0.029). The use of waitlist control in RCTs was associated with significantly better treatment outcomes than the use of other types of control (comparative reduction: −9.27 SUs, 95% CI −13.97 to −4.57, p < 0.001; TR: OR = 3.74, 95% CI 2.13–6.53, p < 0.001). The overall quality of the RCTs was high; a major limitation included high study dropout (43%). Sensitivity analyses confirmed the robustness of our primary analyses. Conclusion To our knowledge, this is the first IPDMA on internet-based interventions that has shown them to be effective in curbing various patterns of adult problem drinking in both community and healthcare settings. Waitlist control may be conducive to inflation of treatment outcomes.
Diese S3-Leitlinie gibt Ärzten und Therapeuten detaillierte Handlungsanweisungen und Therapieempfehlungen für alle alkoholbezogenen Erkrankungen.