Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416

https://doi.org/10.1007/s10488-022-01247-6

ORIGINAL ARTICLE

Programmatic Costs of Project ImPACT for Children with Autism:

A Time‑Driven Activity Based Costing Study
Zuleyha Cidav1,2   · David Mandell1,2 · Brooke Ingersoll3 · Melanie Pellecchia1

Accepted: 20 December 2022 / Published online: 13 January 2023

© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2023

Abstract
Programmatic cost assessment of clinical interventions can inform future dissemination and implementation efforts. We
conducted a randomized trial of Project ImPACT (Improving Parents As Communication Teachers) in which community early
intervention (EI) providers coached caregivers in techniques to improve young children’s social communication skills. We
estimated implementation and intervention costs while demonstrating an application of Time-Driven Activity-Based Costing
(TDABC). We defined Project ImPACT implementation and intervention as processes that can be broken down successively
into a set of procedures. We created process maps for both implementation and intervention delivery. We determined resource
use and costs, per unit procedure in the first year of the program, from a payer perspective. We estimated total implementation
cost per clinician and per site, intervention cost per child, and provided estimates of total hours spent and associated costs for
implementation strategies, intervention activities and their detailed procedures. Total implementation cost was $43,509 per
clinic and $14,503 per clinician. Clinician time (60%) and coach time (12%) were the most expensive personnel resources.
Implementation coordination and monitoring (47%), ongoing consultation (26%) and clinician training (19%) comprised
most of the implementation cost, followed by fidelity assessment (7%), and stakeholder engagement (1%). Per-child inter-
vention costs were $2619 and $9650, respectively, at a dose of one hour per week and four hours per week Project ImPACT.
Clinician and clinic leader time accounted for 98% of per child intervention costs. Highest cost intervention activity was
ImPACT delivery to parents (89%) followed by assessment for child’s ImPACT eligibility (10%). The findings can be used
to inform funding and policy decision-making to enhance early intervention options for young children with autism. Uncom-
pensated time costs of clinicians are large which raises practical and ethical concerns and should be considered in planning
of implementation initiatives. In program budgeting, decisionmakers should anticipate resource needs for coordination and
monitoring activities. TDABC may encourage researchers to assess costs more systematically, relying on process mapping
and gathering prospective data on resource use and costs concurrently with their collection of other trial data.

Keywords  Implementation costs · Intervention costs · Time-driven activity-based costing · Hybrid trial · Autism · Early
intervention

Introduction

Early intervention (EI) services are critical for improving

long-term learning outcomes for children with autism spec-
trum disorder (ASD). Interventions for ASD initiated at ear-
* Zuleyha Cidav lier ages improves children’s cognition, language, behav-
[email protected] ior, and development (Boyd et al., 2010; Bradshaw et al.,
2015; Rogers, 1996). Caregiver-mediated interventions are
1
Department of Psychiatry, Perelman School of Medicine, gaining widespread recognition as an effective and feasible
University of Pennsylvania, Philadelphia, PA, USA
approach to EI for young children with ASD (Green et al.,
2
Leonard Davis Institute of Health Economics, University 2015; Kasari et al., 2014; Pickard et al., 2016). In addition to
of Pennsylvania, Philadelphia, PA, USA
improving child outcomes across a range of developmental
3
Department of Psychology, Michigan State University, domains and reducing challenging behavior, they improve
East Lansing, MI, USA

13
Vol:.(1234567890)
Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416 403

parental self-efficacy, treatment engagement, and stress strategy costs alongside the other rubric elements. The
(Green et al., 2010; Kasari et al., 2014; Rogers et al., 2012; method has been demonstrated with synthetic data generated
Stadnick et al., 2015; Wetherby & Woods, 2006). to exemplify the method but hasn’t been applied in a real-
Project ImPACT (Improving Parents As Communica- world implementation initiative. In addition, the synthetic
tion Teachers) is an evidence-based caregiver coaching example focuses on estimating the implementation strategy
model for families of young children with ASD (Ingersoll costs; the viability of the method for estimating the cost of
& Dvortcsak, 2019). It is a manualized, caregiver-mediated, the intervention being implemented, hasn’t been explored.
naturalistic developmental behavior intervention (Schreib- In this study, we aim to estimate Project ImPACT imple-
man et al., 2015) that includes two core components: (1) a mentation and intervention delivery costs, while demonstrat-
child-directed curriculum to guide caregivers in supporting ing an application of TDABC-Proctor rubric approach in a
their child’s social communication; and (2) guidelines to hybrid trial setting with real-word data. Our cost estimates
help EI providers coach caregivers in using the interven- will be used in a future economic evaluation study to exam-
tion strategies. Project ImPACT has demonstrated efficacy ine intervention outcomes relative to its costs.
in improving caregiver and child outcomes, including
improved caregivers’ treatment adherence and responsive-
ness, and improved children’s language, communication, and Methods
behavior (Stadnick et al., 2015; Stahmer et al., 2020; Yoder
et al., 2021). Project ImPACT​
We collaborated with the Philadelphia Infant Toddler
Early Intervention System (Philadelphia EI system) to Project ImPACT is a parent-mediated intervention for young
conduct a two-year, randomized, implementation-interven- children with autism that blends developmental and natural-
tion (hybrid) trial determining the effectiveness of Project istic behavioral intervention techniques to teach core social
ImPACT in community settings while also examining feasi- communication skills (social engagement, language, social
bility and potential utility of employed implementation strat- imitation, and play) during play and daily routines. To study
egies (hybrid type II trial). (Curran et al., 2012) Philadelphia the effects of Project ImPACT on social communication out-
is one of the first EI systems to implement Project ImPACT comes in children, responsiveness and self-efficacy in their
system wide. Because cost is a leading barrier to adoption parents, as well as the strategies utilized for Project ImPACT
and sustainability of evidence-based practices (EBP) (Aar- implementation, we are conducting a two-year hybrid type
ons et al., 2009; Bond et al., 2014), decision-makers would II (Curran et al., 2012) randomized controlled trial in which
greatly benefit from learning about the budgetary impact of 6 community early intervention agencies in Philadelphia
preparing to implement Project ImPACT in their agencies. are blocked into groups of 3, based on agency size. Within
Uncertainty regarding the costs and cost-effectiveness of each group, each agency was randomized to one of three
such programs can serve as a deterrent to their implementa- conditions: (1) ImPACT at a dose of 1 h per week over six
tion. A comprehensive analysis of programmatic costs can months (low dose Project ImPACT, 24 sessions in total); (2)
mitigate this barrier and provide valuable information for ImPACT at 4 h per week over six months (high dose Project
dissemination efforts by presenting the resource needs and ImPACT, 96 sessions in total); or (3) treatment as usual.
associated costs required for implementation and interven- These agencies are representative of the broader service
tion. Such information is also useful for comparing the costs system regarding the number of clinicians employed (23
of early intervention programs. per agency), and the number of children with or at risk for
Rigorous, detailed, transparent resource use and cost esti- ASD they serve (36 per agency per year). All clinicians in
mates are not typically reported in implementation studies these agencies have a Bachelor’s or Master’s degree in a
(Bowser et al., 2021; Eisman et al., 2020; Gold et al., 2016; relevant field (e.g., psychology, education, speech pathol-
Powell et al., 2019). This is at least partly due to a lack of ogy). Most are independent contractors, provide home-based
clearly defined and standardized costing methods for use in services, use an interdisciplinary approach, and have a treat-
implementation science (Bowser et al., 2021; Dopp et al., ment philosophy that includes caregivers engaging with their
2019a; Roberts et al., 2019). Recently, an approach to cost- children.
ing implementation strategies has been proposed which com- From each of the 6 participating community agencies,
bines Time-Driven Activity-Based Costing (TDABC) with 3 clinicians (18 total) are trained in Project ImPACT via
the Proctor et.al rubric which provides a set of guidelines for a self-paced online tutorial, virtual interactive workshops,
specifying and reporting implementation strategies (Cidav and virtual group and individual consultations. From each
et al., 2020). By blending these two approaches, resource clinician’s caseload, 3 children younger than 30 months of
use and cost estimation is combined with the Proctor rubric age (54 total) and their families are recruited. Although
to allow researchers to routinely estimate implementation Project ImPACT is developed as an in-home intervention,

13

404 Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416

due to Covid-19 challenges, intervention protocol had to be Project ImPACT, introduce it into clinicians’ daily practice,
changed to deliver Project ImPACT virtually over Zoom and establish it in clinics. We define intervention process
platform. Clinicians in the active treatment groups meet as the set of actions that comprise the intervention being
with the families virtually to deliver Project ImPACT every implemented, what the children/families receive as opposed
week for six months either for 1 h or 4 h depending on the to “what otherwise would have happened” (Ovretveit, 2014).
assigned randomization group. Clinicians in the treatment In our specific context, Project ImPACT, implementation
as usual group meet with families virtually for 1 to 3 h each actions are not permanent, are envisioned to be in place only
week to deliver standard early intervention focused on meet- for a certain period (although they might be reintroduced in
ing each child’s individual developmental and communica- the future based on clinic’s specific needs), while interven-
tion goals using play-based therapeutic interactions. tion actions are permanent, and would be in place as long as
Child and parent outcomes are measured through direct the clinic provides Project ImPACT. Hence, implementation
observation and self-report at baseline and 6 months. In costs are one-time startup costs that were incurred in the first
addition to intervention effectiveness, implementation year of program, while intervention costs are ongoing costs.
outcomes will be examined (implementation-intervention Our costing and data collection approach involved the
hybrid type 2 trial) to determine feasibility and potential following:
utility of clinician training and education, ongoing consul-
tation, and fidelity monitoring strategies. Various qualita- Step 1: Name/identify the implementation strategies
tive and quantitative measures on implementation out- and intervention activities and list the associated
comes of acceptability, adoption, feasibility, clinician, and actions, actors, and temporality.
parent fidelity are collected. Analyses of Project ImPACT
implementation and effectiveness outcomes in a publicly For TDABC's process mapping step, we conducted inter-
funded EI setting are ongoing and research publications are views with the key personnel (e.g., study supervisor, project
underway. coordinator) on operational details to fully understand and
document the implementation and intervention process and
TDABC and Proctor et al. Framework for Estimating to create a blueprint of the clinical trial. We identified core
Implementation Strategy and Intervention Costs implementation strategies, intervention activities and their
specific workflows through the study protocol, in close col-
TDABC is based on process mapping which requires sys- laboration with the key study personnel. In naming imple-
tematic, detailed, clear specification of the processes that mentation strategies, we used existing taxonomies in the
can be broken down successively into a set of exact steps to literature (Dopp et al., 2019b; Pinnock et al., 2017; Powell
complete the process. et al., 2015). For each core implementation strategy and
Proctor rubric for implementation strategy specification intervention activity, we delineated specific actions. Then
and reporting (Cidav et al., 2020) suggests when studying for each action, we listed the actors necessary for carrying
implementation strategies, actor (who), action (does what), out a single occurrence of the action and determined their
temporality (when), dosage (frequency and duration), action temporality. In this way, we created a map of the imple-
target, justification and outcomes should be clearly specified. mentation strategies, intervention activities, their associated
Cidav et.al (2020) proposed a 5-step approach to apply the actions, actors, and temporality and validated them with the
TDABC in Implementation Science by conceptualizing an key study personnel.
implementation strategy as a process associated with execut- We defined research activities as activities that would
ing a series of specific procedures performed by using per- not occur typically if the initiative didn’t have an embedded
sonnel resources and non-personnel resources such as equip- research component. As per the input from the EI system
ment and supplies. Proctor rubric is then used for TDABC’s partners and intervention developers, we excluded data col-
process mapping to operationalize implementation strate- lection solely for research purposes, and management, and
gies. This approach can be extended to estimate intervention analysis of these data from process mapping.
costs and to the hybrid trial settings, by conceptualizing the
intervention as a process. Step 2: Determine the frequency and average duration of
each implementation and intervention action by
Data Collection and Cost Estimation actors, and calculate actors’ total time spent on
each action.
We define two overarching processes that comprise: 1. Pro-
ject ImPACT implementation, 2. Project ImPACT interven- For each action, we defined a “time driver” which is
tion. We define implementation process as the set of actions, a feature that would causally affect the time required to
methods taken to enable clinicians and clinics to take up perform the action and measures the volume of the action

13
Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416 405

(e.g. number of consultation sessions). We then deter- Analysis

mined “unit time”, the time required to perform a single
occurrence of the time driver (e.g. duration of a single We conducted a “programmatic (i.e., implementation and
consultation session). We defined the whole process at intervention) cost” study which aims to estimate costs
the time driver- and unit time-levels, allowing each ele- that would fall on those paying for the program; thus, we
ment to be costed on a per instance basis. excluded costs such as broader healthcare system costs or
For all actions, except those related to communica- caregiver time costs. We do include though opportunity cost
tion (meeting, making phone calls, emailing, texting), of clinician time to capture true costs, although payers may
time driver frequencies and unit durations were collected not compensate for these costs.
in real time as part of the trial data collection and their We assessed costs from the payer perspective. The pay-
actual values were documented. For communication ers of the implementation and intervention delivery were,
related actions, real-time documentation was not feasi- respectively, the Eagles Autism Foundation and the Phila-
ble. Meetings were pre-scheduled, recurring, and standard delphia EI system. The time horizon was the first year of
(e.g. monthly meetings) that allowed us to estimate time the study. Since our focus is on estimating Project ImPACT
driver frequencies and durations. For making phone calls, programmatic costs, we only included Project ImPACT par-
emailing, and texting, we relied on vignette-based data ticipants (clinicians and children/caregivers) and excluded
collection (Quigley et al., 2020) by which we observed usual care arm participants. Programmatic costs, as defined
frequencies and unit durations of these actions over a pre- here, are $0 for those who did not participate in the program.
determined period, then extrapolated these to estimate We estimated total implementation cost per clinician and
frequencies and durations over the study period. site, and Project ImPACT intervention cost per child and per
Communication actions also involved communication hour. We provided detailed estimates of total hours spent and
among team members for managing research activities. associated costs by actor, action, implementation strategy
Based on the study team’s input, we assumed that 20% of and intervention activity, and compared them to the total
the communication actions were solely for the research overall cost. When possible, we provided standard deviations
purposes and therefore excluded them from the analysis. for action frequencies and unit durations.
For sensitivity analysis, we estimated costs 1. assum-
Step 3: Determine the price per hour of each actor. ing perfect attendance to Project ImPACT sessions by
caregivers, 2. using United States national average wage
We used actual wage rates of the involved personnel. rates (https://​www.​bls.​gov/) for personnel who are likely
We increased the wage rates by 30% (average employer to carry out the project activities carried out by the study
costs for employee compensation (https://​www.​bls.​gov/)) personnel, 3. including travel to/from child’s home as an
to incorporate workplace benefits, such as health insur- additional intervention action to estimate costs incurred if
ance, paid by the employer. Project ImPACT was delivered in-home, 4. using different
frequencies of actions as they are likely to vary across dif-
Step 4: Determine non-personnel and their associated ferent settings or time periods. We assumed 50% of com-
expenses. munication actions would not occur in the long term when
the intervention matures or in implementation initiatives that
We itemized non-personnel resources and determined do not have an embedded research component, and adjusted
the expense for each item using the project budget and action frequencies accordingly.
expense reports.

Step 5: Calculate total costs.

Results
We calculated actor-level action costs by multiplying
the actor-specific time spent on the action and the actor Detailed estimation of resource use and costs of implementa-
wage rate. We summed action costs at the actor-level to tion, and low dose and high dose intervention can be found
determine implementation strategy and intervention activ- in Tables S1-S3 in the Supplementary File. Tables 1, 2, 3
ity costs at the action-level. The sum of the action costs are simplified versions of these tables. We categorized our
yielded the total personnel time costs. We added non- results by TDABC steps presented above.
personnel costs to obtain the total cost of Project ImPACT
during the study period. Step 1: Name/identify the implementation strategies
and intervention activities and list the associated
actions, actors, and temporality.

13


Table 1  Resource use and costs of project ImPACT implementation strategies*

406

Implementation Strate- Actions Actors Actor hours Wage rate ($) Actor cost ($) Action Action cost ($) Implementation Implementation
gies time strategy time strategy cost ($)

13
(hours) (hours)
A B C D E F G H I J

Stakeholder engagement Meet with clinics indi- Clinic leader 8 82 656 20 1228 31 2027
vidually Project manager 4 81 324
Project coordinator 8 31 248
Meet with clinics as a Clinic leader 8 82 656 11 799
group Project manager 1 81 81
Project coordinator 2 31 62
Clinician training and Conduct self-paced Clinician 72 125 9,000 72 9000 296 32,436
education online training
Conduct interactive Clinician 168 125 21,000 224 23,436
educational meetings Consultant 14 56 784
External trainer 14 56 784
Project coordinator 28 31 868
Ongoing consultation Conduct group consul- Clinician 216 125 27,000 288 30,132 504 45,954
tation sessions Consultant 18 56 1008
External trainer 18 56 1008
Project coordinator 36 31 1116
Conduct individual con- Clinician 54 125 6750 216 15,822
sultation sessions Consultant 162 56 9072
Fidelity Monitoring
Clinician fidelity assess- Watch ImPACT training Clinician 27 125 3375 27 3375 36 3879
ment for ImPACT videos and rate them
direct delivery for fidelity
Evaluate clinicians' Consultant 9 56 504 9 504
fidelity ratings for
ImPACT direct
delivery
Clinician fidelity assess- Watch goal setting ses- Clinician 30 125 3750 60 5430 66 5766
ment for goal setting sion videos and rate Consultant 30 56 1680
them for fidelity
Evaluate clinicians' Consultant 6 56 336 6 336
fidelity ratings for
goal setting
Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416
 Table 1  (continued)
Implementation Strate- Actions Actors Actor hours Wage rate ($) Actor cost ($) Action Action cost ($) Implementation Implementation
gies time strategy time strategy cost ($)
(hours) (hours)
A B C D E F G H I J

Clinician fidelity assess- Watch caregiver coach- Clinician 15 125 1875 30 2715 33 2883
ment for caregiver ing session videos and
coaching rate them for fidelity Consultant 15 56 840

Evaluate clinicians' Consultant 3 56 168 3 168

fidelity ratings for
caregiver coaching
Implementation coordi- Communicate via vir- Clinic leader 40 82 3280 405 26,947 1202 81,093
nation and monitoring tual meetings Clinician 58 125 7250
Consultant 52 56 2912
Project coordinator 114 31 3534
Project manager 83 81 6723
External trainer 58 56 3248
Communicate via adhoc Clinician 36 125 4534 73 6061
phone calls Consultant 16 56 903
Project coordinator 20 31 625
Project manager 0 81 0
External trainer 0 56 0
Communicate via email Clinician 126 125 15,708 537 36,270
Consultant 52 56 2912
Project coordinator 186 31 5776
Project manager 87 81 7020
Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416

External trainer 87 56 4853

Communicate via text Clinician 38 125 4767 187 11,814
Consultant 10 56 582
Project coordinator 95 31 2955
Project manager 43 81 3510
External trainer 0 56 0
Total Hours 2168
Total personnel time 174,038
cost
Non-Personnel Costs 0
Total Implementation 174,038
Cost (1 year, 4 sites,
12 clinicians)

13
407

408 Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416

We defined two phases of the trial, pre-intervention

*Table S1 in Supplementary File provides information on action description, temporality, number of actors in unit action, time driver, time driver frequency, and time driver unit duration, as
strategy cost ($)
Implementation phase and intervention phase, which show the timeline and
flow of the processes. Fig. 1 depicts the process map of the

14,503

43,509
program by phase, delineating implementation strategies,
intervention activities and their specific actions which are
J

listed in the Columns A and B of Tables 1, 2, 3. Action

description and temporality are presented in Tables S1-S3.
Implementation
strategy time

Implementation process consists of 5 discrete implementa-

tion strategies: stakeholder engagement, clinician training
(hours)

and education, ongoing consultation, fidelity assessment

(with 3 discrete sub-strategies: fidelity assessment for Pro-
I

ject ImPACT direct delivery, for goal setting, for parent

Total Implementation

Total Implementation
Cost Per-Clinician

coaching), and implementation coordination and monitor-

ing. Stakeholder engagement involved meeting with clinics
Action cost ($)

Cost Per-Site

individually and as a group. Clinician training and educa-

tion involved conducting self-paced online training and
interactive educational meetings. Ongoing consultation
H

involved conducting individual and group consultation

sessions. Clinician fidelity assessments involved watching
(hours)
Actor hours Wage rate ($) Actor cost ($) Action

recorded videos and rating them for clinician fidelity and

time

evaluating clinicians’ own fidelity ratings. Implementation

G

coordination and monitoring involved communicating via

virtual meetings, ad hoc phone calls, email, and text.
Intervention process consists of 2 activities: assess-
ment for Project ImPACT eligibility and Project ImPACT
F

delivery. Assessment for the Project ImPACT eligibility

involved conducting developmental assessment. Project
ImPACT delivery involved conducting caregiver coaching
sessions, completing administrative work, and conduct-
E

ing case reviews. In sensitivity analyses, this activity also

included travel to/from child’s home.
During the first year of the study, overall personnel con-
sisted of 25 individuals (actors): A team of 1 project man-
D

ager, 2 project coordinators, 1 external trainer, 1 consult-

ant, 12 clinicians, and 8 clinic leaders from 4 clinics. We
report actors on a per action basis that is we listed the job
category and the number of actors within the job category
Actors

involved in a single instance of each action (Tables S1-S3).

For example, in a single occurrence of the interactive edu-
C

cational meeting, 12 clinicians, 1 consultant, 1 external

trainer and 2 project coordinators were present (Table S1).

Step 2: Determine the frequency and average duration of

each implementation and intervention action by
Actions

actors, and calculate actors’ total time spent on

well as details of cost estimation

each action.
B
Implementation Strate-

Column D in Tables 1, 2, 3 shows total time spent on

Table 1  (continued)

each action by each actor. For simplicity, time drivers,

their frequencies and unit durations are presented in Tables
S1-S3.
gies

Step 3: Determine the price per hour of each actor.

A

13
Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416 409

Table 2  Resource use and costs of low dose project ImPACT per-child*

Intervention Actions Actors Actor hours Wage rate ($) Actor cost Action Action cost Activ- Activity cost
Activities ($) time ($) ity time ($)
(hours) (hours)
A B C D E F G H I J

Assess- Conduct Clinician 2 125 250 2 250 2 250

ment for develop-
ImPACT mental-
eligibility behavioral
assessment
Project Conduct Clinician 14 (SD = 7) 125 1750 14 1750 19 2326
ImPACT caregiver
delivery coaching
session
Complete Clinician 4 125 438 4 438
administra-
tive work
Conduct case Clinician 0.67 125 83 1.3 138
review Clinic leader 0.67 82 55
Total hours 21
per child
Personnel 2576
time cost
per child
Non-person- 43
nel costs
per child
(ImPACT
manual)
Total low 2,619
dose
Project
ImPACT
cost per
child
Total low 125
dose
Project
ImPACT
cost per
hour
*
 Low dose: Manualized as 1 h per week for 6 months, 24 sessions in total
Table S2 in Supplementary File provides information on action description, temporality, number of actors in unit action, time driver, time driver
frequency, and time driver unit duration, as well as details of cost estimation

Column E in Tables 1, 2, 3 presents hourly wage rates for through their own devices-computer, phone, iPad. All
clinic leaders, clinicians, consultant, external trainer, project used their own internet service. No technology costs were
coordinator and project manager. incurred in this study.

Step 4: Determine non-personnel and their associated Step 5: Calculate total costs.
expenses.
Action costs by specific actors, total time spent on each
Non-personnel, not time-driven resources shared across action, total cost of an action, total time spent on each imple-
actions included Project ImPACT manual for clinicians and mentation strategy and intervention activity, are given in the
caregivers (Tables 2 and 3). No non-personnel costs were Columns F-J of Tables 1, 2, 3. At the bottom of Tables 1, 2,
incurred other than manuals. Technology used was Zoom 3, we present total personnel and non-personnel costs of the
app which was available to all caregivers and clinicians implementation and intervention costs of Project ImPACT.

13

410 Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416

Table 3  Resource use and costs of high dose project ImPACT per-child*

Intervention Actions Actors Actor hours Wage rate Actor cost Action Action cost Activ- Activity cost
Activities ($) ($) time ($) ity time ($)
(hours) (hours)
A B C D E F G H I J

Assess- Conduct Clinician 2 125 250 2 250 2 250

ment for develop-
ImPACT mental-
eligibility behavioral
assessment
Project Conduct Clinician 59 (SD = 21) 125 7375 59 7375 75 9357
ImPACT caregiver
delivery coaching
session
Complete Clinician 15 125 1844 15 1844
administra-
tive work
Conduct case Clinician 0.67 125 83 1.3 138
review Clinic leader 0.67 82 55
Total hours 77
per child
Personnel 9607
time cost
per child
Non-person- 43
nel costs
per child
(ImPACT
manual)
Total high 9650
dose
Project
ImPACT
cost per
child
Total high 125
dose
Project
ImPACT
cost per
hour
*
 High dose: Manualized as 4 h per week for 6 months, 96 sessions in total
Table S3 in Supplementary File provides information on action description, temporality, number of actors in unit action, time driver, time driver
frequency, and time driver unit duration, as well as details of cost estimation

In the first year of the program, estimated total personnel (19%) represented most of the implementation cost, followed
hours devoted to implementation was 2168 with associated by fidelity assessment (7%), and stakeholder engagement
total implementation cost of $174,038 (2021 price year), (1%). Highest cost implementation actions were communi-
$43,509 per clinic, $14,503 per clinician, of which all were cation via email exchanges (21%), phone/virtual meetings
personnel time costs (Table 1). Of these costs, only 40% (15%), followed by conducting group consultation (17%),
were compensated by the Philadelphia EI system. The rest interactive educational meetings (13%), individual consulta-
constituted uncompensated clinician time cost (60% of total tion (9%), and communicating via text (7%). The rest of the
costs). Consultant time (12%) was the next most expensive implementation action costs were relatively small (< 3%).
personnel resource, followed by project manager (10%), Implementation costs did not vary by low dose vs high dose
project coordinator (9%), external trainer (6%) and clinic intervention arms, since all clinicians were subject to the
leader (3%). Implementation coordination and monitoring same implementation tasks as per Project ImPACT training
(47%), ongoing consultation (26%) and clinician training protocol regardless of their assignment to intervention arms.

13
Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416 411

Fig. 1  Process map of implementation and delivery of the project ImPACT​

For virtually conducted, low dose Project ImPACT, deliv- and duration and had to be completed by each clinician for
ered over six months, per child intervention cost was $2576 them to be able to get certification.
(21 h of personnel time, $125 per hour) of which 1% was For intervention activities, developmental assessment,
expense for Project ImPACT manuals (Table 2), and the administrative work, and case review were very standard
remaining was cost of clinician (98%) and clinic leader time actions, their frequencies and unit durations did not vary
(1%). For virtually delivered, high dose Project ImPACT meaningfully across or within clinicians or clinic leaders.
delivered over six months, per child cost was $9650 (77 h of Caregiver coaching sessions are designed and manualized as
personnel time, $125 per hour) (Table 3). Highest cost inter- 1 h sessions. They were scheduled for an hour with caregiv-
vention activity was Project ImPACT delivery to caregivers ers; there was not much room for flexibility in duration since
(89%) followed by assessment of child’s ImPACT eligibility the providers were clinicians with large workloads, mostly
(10%). Highest cost ImPACT delivery actions were conduct- contract based, who had other appointments to attend. Num-
ing caregiver coaching sessions (68%), completing adminis- ber of caregiver coaching sessions per child, however, varied
trative work (17%), followed by conducting developmental across clinicians. Number of coaching sessions per child for
assessments (10%) and conducting case reviews (5%). low dose Project ImPACT was 14 (SD = 7), and 59 (SD = 21)
Regarding the variance estimates, cost variation could for high dose Project ImPACT.
occur due to variations in activity frequencies and unit dura-
tions across or within the actors in the same job category who
perform a specific task. In most cases, there was not sufficient Sensitivity Analysis
variation due to low sample size of actors (1 project manager, 2
project coordinators, 1 external trainer, 1 consultant). In addi- Assuming perfect attendance to Project ImPACT coaching
tion, for clinicians, implementation actions were protocolized sessions by caregivers, per child intervention cost was $4181
for Project ImPACT certification purposes. Same implementa- for low dose Project ImPACT and $15,431 for high dose
tion tasks were assigned to all clinicians at the same frequency Project ImPACT.

13

412 Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416

Using United States national average wage rates for intervention might have to be tailored or adapted to meet the
personnel who are likely to carry out the project activi- community’s needs and available resources.
ties carried out by the study personnel (social and com- Reporting resource use in units and their unit prices, as
munity service manager for clinic leader:$47, therapist opposed to reporting lumpsum total $ amounts, is helpful to
for clinician:$39, psychologist for consultant and external circumvent the challenge of generalizability of cost estimates
trainer:$62, administrative support worker for project coor- to other settings. When costs are disaggregated and reported
dinator:$26, medical, health services manager for project in this way, a user of this data interested in replicating an
manager $74), (https://​www.​bls.​gov/) first year, one-time intervention in their local setting could apply unit prices
implementation cost was $8270 per clinician and $24,810 specific to their setting, change resource units according to
per site. Using national average wage rates, per child inter- their unique needs (e.g., more clinician hours are needed
vention cost was $842 for low dose intervention and $3055 due to specific challenges in their setting), change a specific
for high dose intervention. component of the implementation strategy (eg. provide only
If we Include travel to/from child’s home as an additional didactic training, drop ongoing consultation as it fits to their
intervention action to estimate costs incurred if Project context) and see cost implications of these changes.
ImPACT was delivered in-home, per child cost would be Our implementation cost estimates account for the oppor-
$4369 for low dose intervention and $17,025 for high dose tunity costs, the efforts and time investments by the com-
intervention. munity partners for which they may not be compensated,
When we decreased the frequency of communication and therefore represent true economic costs. These costs are
actions by 50%, the first year, one-time implementation cost not trivial; indeed, of the twelve clinicians, ten were con-
was $11,520 per clinician and $34,561 per site. tract employees who were not compensated for their time
spent on training, consultation, and other implementation
activities. Cost of their uncompensated time constituted the
Discusson largest cost of implementation. Uncompensated labor of
public mental health clinicians exacerbates economic pre-
Estimating costs of implementing and delivering EI pro- carity, financial strain and job-related stress among public
grams supports policy makers and health administrators in mental health clinicians (Last et al., 2022). This also puts a
making budgeting decisions and allocating scarce healthcare burden on the publicly funded EI system and the EI agen-
resources efficiently. In this study, we estimated implemen- cies who are themselves facing financial constraints. These
tation and intervention delivery costs of Project ImPACT costs should be routinely captured in costing studies and
while demonstrating an application of TDABC-Proctor considered in planning of the implementation initiatives
rubric approach in a hybrid trial setting with real-word data. and developing policies to support and retain public mental
In the first year of the program, estimated total personnel health workforce.
hours devoted to implementation was 2168 with associated It is difficult to compare our implementation cost esti-
total implementation cost of $174,038 (2021 price year), mates to that of other similar parent-mediated programs
$43,509 per clinic, $14,503 per clinician, of which all were due to scarcity of cost studies in this area. Our per clinician
personnel time costs. For low dose Project ImPACT, per implementation cost estimate of $14,503 is almost as twice
child intervention cost was $2576 (21 h of personnel time) as what has been reported for per therapist cost for the Par-
of which 1% was expense for Project ImPACT manuals, and ent–Child Interaction Therapy (PCIT) training and consulta-
the remaining was time cost. For high dose Project ImPACT, tion in the Philadelphia behavioral health system (Okamura
per child cost was $9650 (77 h of personnel time). et al., 2017). However, these costs are not directly compara-
Our costing approach yielded useful information about ble due to methodological differences in cost estimation. In
what programmatic factors have important impact on imple- future studies, transparent and detailed description of cost-
mentation and intervention costs. In this case, per clinician ing methods used, and detailed activity level resource use
and per child costs are affected by (1) specific activities and costs would facilitate cross study comparisons and the
performed, (2) frequency of activities, (3) average duration potential for the findings to drive the policy decisions.
per activity, (4) personnel involved in activities, and (5) Implementation coordination and monitoring costs that
personnel wage rates. Any change in any of these factors arose from email and text correspondences and virtual/
will have cost implications. An inherent challenge in rand- phone meetings constituted almost half of the implementa-
omized controlled trials is generalizability of findings (not tion costs. High administrative costs have been observed in
only cost-related findings, but also effectiveness results) to other studies that emphasize the importance of communi-
other settings. In implementation settings that do not have an cation costs (Cidav et al., 2021; Ingels et al., 2016; Smith
embedded research component, intervention might be car- et al., 2020; Subramanian et al., 2011). Although we aimed
ried out differently than in controlled trials. For example, to account for it in our sensitivity analysis, these large costs

13
Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416 413

might still be due in part to the fact that this analysis took research comparing in-home vs. telehealth delivery of Pro-
place alongside a research study and therefore study per- ject ImPACT found no differences in parent or child out-
sonnel made greater than average communication efforts to comes (Hao et al., 2021), suggesting that virtual delivery
ensure that trial activities are performed as per protocol. may be a cost-effective option. To explore this possibility,
Nevertheless, in budgeting for future programs, decision- future studies should evaluate cost effectiveness of different
makers should anticipate resource needs related to coordina- delivery modes.
tion and monitoring activities and assess how these activities This study has several important strengths. To our knowl-
can be performed most efficiently (Dopp et al., 2020, 2021). edge, it is the first to examine implementation resources and
As programs mature the proportion of total costs for such costs of a manualized EI program in a community setting,
activities may decrease because of more efficient program while also providing resource use and cost estimates for
management and the routinization of certain procedures. the intervention itself. We demonstrated an application of
Project ImPACT intervention or ongoing cost was low TDABC alongside Proctor rubric, in a hybrid trial setting
relative to implementation or startup cost. Many studies have to estimate both implementation and intervention costs. We
noted substantial start-up costs that are separate from the carefully described our methods, including the TDABC
ongoing costs of delivering the interventions themselves. setup, data collection, and assigning a dollar value to the
(Bowser et  al., 2021) In addition to informing program resources used, and provided a transparent and detailed com-
design, these cost estimates may be useful for comparative position of resource use and costs across various, granular
purposes when considering the costs of other EI programs. implementation and intervention tasks. This level of detail
The costs of EI programs for ASD are substantial; most US is often missing, making it difficult to compare the results of
studies indicate that the annual cost for such programs range the studies and evaluate the quality of the cost collection and
between $40,000 and $80,000 per child (Rogge & Jans- valuation (Eisman et al., 2021; Gold et al., 2022). Having
sen, 2019). With $125 per hour cost, if high dose Project disaggregated, detailed information helps decision makers
ImPACT were to be delivered for 52 weeks, totaling 208 h, understand the true cost to their organizations of implement-
annual cost of Project ImPACT delivery would be $26,000. ing and delivering Project ImPACT.
One should always be careful when comparing such esti- Another strength is the prospective estimation of costs
mates. In most cases, direct comparison is not possible due that did not rely on retrospective self-reports, avoiding
to differences in methodological approach and underlying potential recall bias. Once Project ImPACT implementa-
assumptions. For accurate cost comparisons across studies, tion and intervention processes were mapped, process data
standardized methods of costing and transparency in pre- (e.g., number of group consultation sessions, duration of
senting the cost analysis are essential. Future cost studies one group session) were captured in real-time from the trial
conducted prospectively to capture detailed activity level documentation. This minimized the data collection burden
costs would make such meaningful comparisons possible. on study personnel. Previous TDABC applications have
Given, previous evidence on Project ImPACT’s efficacy demonstrated that determining activity frequencies and
(Stadnick et al., 2015; Stahmer et al., 2020; Yoder et al., establishing a unit duration are less burdensome than asking
2021) and our cost estimates, it might be possible to pro- personnel to log their time and activities (Kaplan & Ander-
vide high quality EI for young children with ASD at sig- son, 2007; Keel et al., 2017; Quigley et al., 2020; Silva Etges
nificantly lower costs. However, more evidence on Project et al., 2019). Only for as- needed or ad hoc communication
ImPACT’s effectiveness on child outcomes and its costs in (e.g., phone calls, emails) activities, we needed to rely on
different settings and populations are warranted to explore vignette-based estimation in which staff tracked their ad hoc
this possibility. communication frequency and duration during two typical
We found significant travel costs associated with in-home work weeks. Nonetheless, the majority of implementation
delivery of Project ImPACT. We estimated that the cost of and intervention actions were relatively straightforward to
delivering Project ImPACT in-home would be almost twice capture using this method.
as much as the cost of delivering it virtually. This is in line Another methodological strength is that this costing
with the findings of the previous literature that found an approach allows estimation of replication costs that would be
outpatient model is 2.62 times more expensive, and an in- incurred in other settings or under different conditions, that
home model is 2.64 times more expensive than telehealth is when the implementation strategies or Project ImPACT
(Little et al., 2018). Early intervention systems may be able intervention must be tailored or adapted to meet the com-
to increase the number of families that they serve, especially munity’s needs and available resources. For example, the
in rural and underserved areas, and decrease the program number of individual consultation sessions may have to be
costs by using such innovative models of service delivery, reduced in low resource settings; replacing the frequency of
such as combining face-to-face interactions with telehealth individual consultation sessions in Table S1 and recalculat-
sessions. (Little et al., 2018; Pickard et al., 2016) Recent ing the total costs would address this situation. If a provider

13

414 Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416

expects to pay a different salary to consultants and external data about its effectiveness in community settings are cur-
trainers, an updated hourly rate would be multiplied by the rently being collected. Combining cost data with forthcom-
time provided in Table S1 to determine an estimated total ing outcome data to conduct economic evaluations should
personnel cost. be useful to decisionmakers working with scarce resources.
There are limitations to this study. First, the study was Use of TDABC with the Proctor rubric may encourage
conducted as part of a randomized field trial, so the results implementation researchers to perform costing studies more
may not be applicable to non-trial settings. However, we regularly, relying on process mapping and collecting pro-
provided sufficient information to estimate replication costs spective data on costs concurrently with other trial data col-
as described above. Second, implementation coordination lection. This may enhance the knowledge base in this area
and monitoring costs may have been overestimated because with standardized, detailed, transparent, and quality cost
some communication may have only occurred because they information and evaluation. Results of the clinical effective-
were part of the study, although we tried to account for this ness of a program, can then be simultaneously presented
in our sensitivity analysis. However, given the detailed, along with program cost findings. This would inform future
granular information, these costs can easily be recalculated replication, dissemination, adoption, implementation, and
for varying communication action frequency and unit dura- economic evaluation of new interventions and contribute
tions. Third, frequency and unit duration of communica- to methodological advances in Implementation Science
tion activities were derived from vignette-based estimations regarding standardized methods for tracking and reporting
because these were not collected as part of other trial data of implementation strategies.
collection, and it was not possible for personnel to record
Supplementary Information  The online version contains supplemen-
each instance of communication actions. Fourth, insuf- tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 10488-0​ 22-0​ 1247-6.
ficient variation in our data precluded us from estimating
variation in cost estimates and examining uncertainty with
a probabilistic costing approach. Lastly, we performed this
analysis from a programmatic perspective. A more com- References
prehensive cost estimation from a wider perspective would
include time costs of the caregivers participating in Pro- Aarons, G. A., Wells, R. S., Zagursky, K., Fettes, D. L., & Palinkas, L.
A. (2009). Implementing evidence-based practice in community
ject ImPACT (e.g., time spent participating in the program
mental health agencies: A multiple stakeholder analysis. American
activities) and costs associated with the changes in children’s Journal of Public Health, 99, 2087–2095.
use of other services. In addition, we did not provide a full Bond, G. R., Drake, R. E., McHugo, G. J., Peterson, A. E., Jones, A.
economic evaluation in which outcomes and costs observed M., & Williams, J. (2014). Long-term sustainability of evidence-
based practices in community mental health agencies. Administra-
for the intervention arm are compared to those of usual care
tion and Policy in Mental Health, 41, 228–236.
arm. Establishing the economic value of Project ImPACT Bowser, D. M., Henry, B. F., & McCollister, K. E. (2021). Cost analysis
requires examining its benefits in relation to its costs; our in implementation studies of evidence-based practices for mental
cost estimates provide the first step in determining Project health and substance use disorders: A systematic review. Imple-
mentation Science, 16, 26.
ImPACT’s economic value. Current studies are underway
Boyd, B. A., Odom, S. L., Humphreys, B. P., & Sam, A. M. (2010).
to broaden our cost analysis perspective to account for the Infants and toddlers with autism spectrum disorder: Early iden-
possible changes in children’s’ use of healthcare services and tification and early intervention. Journal of Early Intervention.,
assess the program’s cost-effectiveness. 32, 75–98.
Bradshaw, J., Steiner, A. M., Gengoux, G., & Koegel, L. K. (2015).
Feasibility and effectiveness of very early intervention for infants
at-risk for autism spectrum disorder: A systematic review. Journal
Conclusions of Autism and Developmental Disorders, 45, 778–794.
Cidav, Z., Mandell, D., Pyne, J., Beidas, R., Curran, G., & Marcus,
S. (2020). A pragmatic method for costing implementation strat-
Our resource use and cost estimates can serve as a reference
egies using time-driven activity-based costing. Implementation
point to publicly funded EI systems who may wish to adopt Science, 15, 28.
Project ImPACT. The specific implementation and interven- Cidav, Z., Marcus, S., Mandell, D., et al. (2021). Programmatic costs of
tion procedures and staffing arrangements we identified in the telehealth ostomy self-management training: An application of
time-driven activity-based costing. Value Health., 24, 1245–1253.
this study can function as a model to implement the pro-
Curran, G. M., Bauer, M., Mittman, B., Pyne, J. M., & Stetler, C.
gram. However, cost is only one piece of information needed (2012). Effectiveness-implementation hybrid designs: Combining
for making decisions about community-based implementa- elements of clinical effectiveness and implementation research to
tion of Project ImPACT; outcome information also is nec- enhance public health impact. Medical Care, 50, 217–226.
da Silva Etges, A. P. B., Cruz, L. N., Notti, R. K., et al. (2019). An
essary. Efficacy evaluation of Project ImPACT suggests the
8-step framework for implementing time-driven activity-based
program improves caregiver and child outcomes (Stadnick costing in healthcare studies. The European Journal of Health
et al., 2015; Stahmer et al., 2020; Yoder et al., 2021), and Economics, 20, 1133–1145.

13
Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416 415

Dopp, A. R., Kerns, S. E. U., Panattoni, L., et al. (2021). Translating Okamura, K. H., Benjamin Wolk, C. L., Kang-Yi, C. D., et al. (2017).
economic evaluations into financing strategies for implement- The price per prospective consumer of providing therapist train-
ing evidence-based practices. Implementation Science, 16, 66. ing and consultation in seven evidence-based treatments within
Dopp, A. R., Mundey, P., Beasley, L. O., Silovsky, J. F., & Eisenberg, a large public behavioral health system: An example cost-anal-
D. (2019a). Mixed-method approaches to strengthen economic ysis metric. Frontiers in Public Health, 5, 356.
evaluations in implementation research. Implementation Sci- Ovretveit, J. (2014). Evaluating improvement and implementation
ence, 14, 2. for health. McGraw-Hill Education (UK).
Dopp, A. R., Parisi, K. E., Munson, S. A., & Lyon, A. R. (2019b). A Pickard, K. E., Wainer, A. L., Bailey, K. M., & Ingersoll, B. R.
glossary of user-centered design strategies for implementation (2016). A mixed-method evaluation of the feasibility and
experts. Transl Behav Med., 9, 1057–1064. acceptability of a telehealth-based parent-mediated interven-
Dopp, A. R., Narcisse, M.-R., Mundey, P., et al. (2020). A scop- tion for children with autism spectrum disorder. Autism, 20,
ing review of strategies for financing the implementation of 845–855.
evidence-based practices in behavioral health systems: State of Pinnock, H., Barwick, M., Carpenter, C. R., et al. (2017). Standards
the literature and future directions. Implementation Research for Reporting Implementation Studies (StaRI) Statement. BMJ,
and Practice., 1, 2633489520939980. 356, i6795.
Eisman, A. B., Kilbourne, A. M., Dopp, A. R., Saldana, L., & Eisen- Powell, B. J., Fernandez, M. E., Williams, N. J., et  al. (2019).
berg, D. (2020). Economic evaluation in implementation sci- Enhancing the Impact of Implementation Strategies in Health-
ence: Making the business case for implementation strategies. care: A Research Agenda. Frontiers in Public Health, 7, 3.
Psychiatry Research, 283, 112433. Powell, B. J., Waltz, T. J., Chinman, M. J., et al. (2015). A refined
Eisman, A. B., Quanbeck, A., Bounthavong, M., Panattoni, L., & compilation of implementation strategies: Results from the
Glasgow, R. E. (2021). Implementation science issues in under- Expert Recommendations for Implementing Change (ERIC)
standing, collecting, and using cost estimates: a multi-stake- project. Implementation Science, 10, 21.
holder perspective. Implement Science, 16(1), 75. Quigley, E., O’Donnell, S., & Doyle, G. (2020). Time-Driven Activ-
Gold, R., Bunce, A. E., Cohen, D. J., et al. (2016). Reporting on the ity-Based Costing: A Step-by-Step Guide to Collecting Patient-
strategies needed to implement proven interventions: An exam- Level Health Care Costs. SAGE Publications Ltd.
ple from a “Real-World” cross-setting implementation study. Roberts, S. L. E., Healey, A., & Sevdalis, N. (2019). Use of health
Mayo Clinic Proceedings, 91, 1074–1083. economic evaluation in the implementation and improvement
Gold, H. T., McDermott, C., Hoomans, T., & Wagner, T. H. (2022). science fields-a systematic literature review. Implementation
Cost data in implementation science: Categories and approaches Science, 14, 72.
to costing. Implementation Science, 17, 11. Rogers, S. J. (1996). Brief report: Early intervention in autism. Jour-
Green, J., Charman, T., McConachie, H., et  al. (2010). Parent- nal of Autism and Developmental Disorders, 26, 243–246.
mediated communication-focused treatment in children with Rogers, S. J., Estes, A., Lord, C., et al. (2012). Effects of a brief
autism (PACT): A randomised controlled trial. Lancet, 375, Early Start Denver model (ESDM)-based parent intervention
2152–2160. on toddlers at risk for autism spectrum disorders: A randomized
Green, J., Charman, T., Pickles, A., et al. (2015). Parent-mediated controlled trial. Journal of the American Academy of Child and
intervention versus no intervention for infants at high risk of Adolescent Psychiatry, 51, 1052–1065.
autism: A parallel, single-blind, randomised trial. Lancet Psy- Rogge, N., & Janssen, J. (2019). The economic costs of autism spec-
chiatry, 2, 133–140. trum disorder: A literature review. Journal of Autism and Devel-
Hao, Y., Franco, J. H., Sundarrajan, M., & Chen, Y. (2021). A pilot opmental Disorders, 49, 2873–2900.
study comparing tele-therapy and in-person Therapy: Perspec- Schreibman, L., Dawson, G., Stahmer, A. C., et al. (2015). Natu-
tives from parent-mediated intervention for children with autism ralistic developmental behavioral interventions: Empirically
spectrum disorders. Journal of Autism and Developmental Dis- validated treatments for autism spectrum disorder. Journal of
orders, 51, 129–143. Autism and Developmental Disorders, 45, 2411–2428.
Ingels, J. B., Walcott, R. L., Wilson, M. G., et al. (2016). A Prospec- Smith, M. J., Graham, A. K., Sax, R., et al. (2020). Costs of pre-
tive programmatic cost analysis of fuel your life: A worksite paring to implement a virtual reality job interview training
translation of DPP. Journal of Occupational and Environmental programme in a community mental health agency: A budget
Medicine, 58, 1106–1112. impact analysis. Journal of Evaluation in Clinical Practice, 26,
Ingersoll, B., & Dvortcsak, A. (2019). Teaching Social Communica- 1188–1195.
tion to Children with Autism (p. 386). Guilford Press. Stadnick, N. A., Stahmer, A., & Brookman-Frazee, L. (2015). Pre-
Kaplan, R. S., & Anderson, S. R. (2007). Time-driven activity-based liminary effectiveness of project ImPACT: A parent-mediated
costing (p. 266). Harvard Business Press. intervention for children with autism spectrum disorder deliv-
Kasari, C., Lawton, K., Shih, W., et al. (2014). Caregiver-mediated ered in a community program. Journal of Autism and Develop-
intervention for low-resourced preschoolers with autism: An mental Disorders, 45, 2092–2104.
RCT. Pediatrics, 134, e72–e79. Stahmer, A. C., Rieth, S. R., Dickson, K. S., et al. (2020). Project
Keel, G., Savage, C., Rafiq, M., & Mazzocato, P. (2017). Time- ImPACT for Toddlers: Pilot outcomes of a community adapta-
driven activity-based costing in health care: A systematic tion of an intervention for autism risk. Autism, 24, 617–632.
review of the literature. Health Policy, 121, 755–763. Subramanian, S., Tangka, F. K., Hoover, S., Degroff, A., Royalty,
Last, B. S., Schriger, S. H., Becker-Haimes, E. M., et al. (2022). Eco- J., & Seeff, L. C. (2011). Clinical and programmatic costs of
nomic precarity, financial strain, and job-related stress among implementing colorectal cancer screening: Evaluation of five
Philadelphia’s public mental health clinicians. Psychiatric Ser- programs. Evaluation and Program Planning, 34, 147–153.
vices (washington, d. c.), 73, 774–786. U.S. Bureau of Labor Statistics. https://​www.​bls.​gov/ Last Accessed
Little, L. M., Wallisch, A., Pope, E., & Dunn, W. (2018). Accept- October 12.
ability and cost comparison of a telehealth intervention for Wetherby, A. M., & Woods, J. J. (2006). Early social interaction
families of children with autism. Infants & Young Children., project for children with autism spectrum disorders beginning
31, 275–286. in the second year of life: A preliminary study. Topics Early
Childhood Special Education., 26, 67–82.

13

416 Administration and Policy in Mental Health and Mental Health Services Research (2023) 50:402–416

Yoder, P. J., Stone, W. L., & Edmunds, S. R. (2021). Parent utiliza- Springer Nature or its licensor (e.g. a society or other partner) holds
tion of ImPACT intervention strategies is a mediator of proxi- exclusive rights to this article under a publishing agreement with the
mal then distal social communication outcomes in younger author(s) or other rightsholder(s); author self-archiving of the accepted
siblings of children with ASD. Autism, 25, 44–57. manuscript version of this article is solely governed by the terms of
such publishing agreement and applicable law.
Publisher's Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.

13