Advertisement

Predicting and responding to change: Perceived environmental uncertainty among substance use disorder treatment programs

Published:January 06, 2023DOI:https://doi.org/10.1016/j.josat.2022.208947

      Highlights

      • Treatment programs must be prepared to predict and respond to uncertainties in their environment.
      • Reported difficulty predicting and responding to changes decreased from 2014 to 2017, but remained relatively high by 2017.
      • Multiple levels of characteristics are associated with difficulty in predicting and responding to, and effect of change.
      • Interventions that facilitate effective prediction and response to change are needed to improve care delivery and outcomes.

      Abstract

      Introduction

      Substance use disorder (SUD) treatment programs offering addiction health services (AHS) must be prepared to adapt to change in their operating environment. These environmental uncertainties may have implications for service delivery, and ultimately patient outcomes. To adapt to a multitude of environmental uncertainties, treatment programs must be prepared to predict and respond to change. Yet, research on treatment programs preparedness for change is sparse. We examined reported difficulties in predicting and responding to changes in the AHS system, and factors associated with these outcomes.

      Methods

      Cross-sectional surveys of SUD treatment programs in the United States in 2014 and 2017. We used linear and ordered logistic regression to examine associations between key independent variables (e.g., program, staff, and client characteristics) and four outcomes, (1) reported difficulties in predicting change, (2) predicting effect of change on organization, (3) responding to change, and (4) predicting changes to make to respond to environmental uncertainties. Data were collected through telephone surveys.

      Results

      The proportion of SUD treatment programs reporting difficulty predicting and responding to changes in the AHS system decreased from 2014 to 2017. However, a considerable proportion still reported difficulty in 2017. We identified that different organizational characteristics are associated with their reported ability to predict or respond to environmental uncertainty. Findings show that predicting change is significantly associated with program characteristics only, while predicting effect of change on organizations is associated with program and staff characteristics. Deciding how to respond to change is associated with program, staff, and client characteristics, while predicting changes to make to respond is associated with staff characteristics only.

      Conclusions

      Although treatment programs reported decreased difficulty predicting and responding to changes, our findings identify program characteristics and attributes that could better position programs with the foresight to more effectively predict and respond to uncertainties. Given resource constraints at multiple levels in treatment programs, this knowledge might help identify and optimize aspects of programs to intervene upon to enhance their adaptability to change. These efforts may positively influences processes or care delivery, and ultimately translate into improvements in patient outcomes.

      Keywords

      1. Introduction

      Organizations providing addiction health services (AHS) must contend with considerable uncertainty in the course of service delivery, including anticipating and responding to change in the substance use disorder (SUD) treatment environment. The AHS system, in particular, has faced significant policy changes that have increased the level of uncertainty in aspects of their operations, including payment and service delivery (
      • Carlo A.D.
      • Benson N.M.
      • Chu F.
      • Busch A.B.
      Association of Alternative Payment and Delivery Models with Outcomes for mental health and substance use disorders: A systematic review.
      ). These organizations face the challenge of adapting to a multitude of fundamental industry changes. New public policies and changes in market demands have called upon healthcare organizations to integrate previously-specialized functions, such as emergency preparedness (
      • Lurie N.
      • Wasserman J.
      • Nelson C.D.
      Public health preparedness: Evolution or revolution?.
      ); adapt to technological and operational innovations (
      • Substance Abuse and Mental Health Services Administration a.
      Office of the Surgeon General
      Chapter 6: Health care systems and substance use disorders.
      ;
      • Tai B.
      • Volkow N.D.
      Treatment for substance use disorder: Opportunities and challenges under the affordable care act.
      ); and reduce healthcare disparities through practices, such as reporting demographic data and hiring more diverse staff (
      • Guerrero E.G.
      • Khachikian T.
      • Frimpong J.A.
      • Kong Y.
      • Howard D.L.
      • Hunter S.
      Drivers of continued implementation of cultural competence in substance use disorder treatment.
      ). Policies such as The Affordable Care Act (ACA) also promoted a number of changes in the health care environment, including coordination of care, improving reliability, and lowering prices, which may in turn influence how programs predict or respond to change (
      • McAlearney A.S.
      • Walker D.M.
      • Hefner J.L.
      Moving organizational culture from volume to value: A qualitative analysis of private sector accountable care organization development.
      ;
      • Polonsky M.S.
      High-reliability organizations: The next frontier in healthcare quality and safety.
      ).
      A wide range of sources have been used to represent uncertainty in the literature, including complexity and number of factors involved, frequency of change in the organization's operating environments, and insufficient information (
      • Duncan R.B.
      Characteristics of organizational environments and perceived environmental uncertainty.
      ;
      • Han P.K.J.
      • Klein W.M.P.
      • Arora N.K.
      Varieties of uncertainty in health care: A conceptual taxonomy.
      ;
      • Kreiser P.
      • Marino L.
      Analyzing the historical development of the environmental uncertainty construct.
      ). One of the most impactful ways that organizations mitigate uncertainty to improve their ability to predict and respond to changes is through the use of computer-based technologies such as electronic health records (EHR). Healthcare information systems have been used to reduce uncertainty and improve decision-making by allowing healthcare organizations to monitor key operational factors, quickly access medical records, identify patterns, and model the effects of potential courses of action (
      • Callahan A.
      • Shah N.H.
      Chapter 19—Machine learning in healthcare.
      ;
      • Frimpong J.A.
      • Jackson B.E.
      • Stewart L.M.
      • Singh K.P.
      • Rivers P.A.
      • Bae S.
      Health information technology capacity at federally qualified health centers: A mechanism for improving quality of care.
      ;
      • Hamrock E.
      • Paige K.
      • Parks J.
      • Scheulen J.
      • Levin S.
      Discrete event simulation for healthcare organizations: A tool for decision making.
      ;
      • Lee S.
      • Song J.
      • Cao Q.
      Environmental uncertainty and firm performance: An empirical study with strategic alignment in the healthcare industry. ICIS 2011 proceedings.
      ;
      • Norgeot B.
      • Glicksberg B.S.
      • Butte A.J.
      A call for deep-learning healthcare.
      ;
      • Thakur R.
      • Hsu S.H.Y.
      • Fontenot G.
      Innovation in healthcare: Issues and future trends.
      ). In addition, organizations report better adaptation to change when employees are given a meaningful role in designing and enacting the response (
      • Longenecker C.O.
      • Longenecker P.D.
      Why hospital improvement efforts fail: A view from the front line.
      ;
      • Thakur R.
      • Hsu S.H.Y.
      • Fontenot G.
      Innovation in healthcare: Issues and future trends.
      ). Limited evidence suggests that higher program age is also positively associated with adoption of innovative SUD treatment practices (
      • Roman P.M.
      • Johnson J.A.
      Adoption and implementation of new technologies in substance abuse treatment.
      ).
      Managerial characteristics may also affect a health care organization's ability to predict and respond to changes. The results of existing studies on the association between medical personnel’s demographic characteristics (e.g., race, sex, years of training) and their ability to tolerate uncertainty have been mixed (
      • Strout T.D.
      • Hillen M.
      • Gutheil C.
      • Anderson E.
      • Hutchinson R.
      • Ward H.
      • Kay H.
      • Mills G.J.
      • Han P.K.J.
      Tolerance of uncertainty: A systematic review of health and healthcare-related outcomes.
      ). Decisive and transformational leaders are well recognized as important factors in successful navigation of change in various organizational settings (
      • Faupel S.
      • Süß S.
      The effect of transformational leadership on employees during organizational change – an empirical analysis.
      ;
      • Green C.A.
      • McCarty D.
      • Mertens J.
      • Lynch F.L.
      • Hilde A.
      • Firemark A.
      • Weisner C.M.
      • Pating D.
      • Anderson B.M.
      A qualitative study of the adoption of buprenorphine for opioid addiction treatment.
      ;
      • Guerrero E.G.
      • Khachikian T.
      • Frimpong J.A.
      • Kong Y.
      • Howard D.L.
      • Hunter S.
      Drivers of continued implementation of cultural competence in substance use disorder treatment.
      ;
      • Lurie N.
      • Wasserman J.
      • Nelson C.D.
      Public health preparedness: Evolution or revolution?.
      ). Studies on SUD treatment organizations have identified additional change-facilitating leadership characteristics, including longer tenure (
      • Friedmann P.D.
      • Jiang L.
      • Alexander J.A.
      Top manager effects on buprenorphine adoption in outpatient substance abuse treatment programs.
      ;
      • Roman P.M.
      • Johnson J.A.
      Adoption and implementation of new technologies in substance abuse treatment.
      ) and larger information networks through greater involvement in professional associations (
      • Ducharme L.J.
      • Knudsen H.K.
      • Roman P.M.
      • Johnson J.A.
      Innovation adoption in substance abuse treatment: Exposure, trialability, and the clinical trials network.
      ;
      • Fields D.
      • Knudsen H.K.
      • Roman P.M.
      Implementation of network for the improvement of addiction treatment (NIATx) processes in substance use disorder treatment centers.
      ;
      • Savage S.A.
      • Abraham A.J.
      • Knudsen H.K.
      • Rothrauff T.C.
      • Roman P.M.
      Timing of buprenorphine adoption by privately funded substance abuse treatment programs: The role of institutional and resource-based inter-organizational linkages.
      ).
      Structural factors, including ownership type and for- vs. non-profit status, are also associated with preparedness to adapt to change in health care organizations (
      • Freedman S.
      • Lin H.
      Hospital ownership type and innovation: The case of electronic medical records adoption.
      ;
      • Shields M.C.
      • Horgan C.M.
      • Ritter G.A.
      • Busch A.B.
      Use of electronic health information Technology in a National Sample of hospitals that provide specialty substance use care.
      ).
      In the context of organizational management literature, the “uncertainty” that organizations face with regards to predicting and responding to changes has been widely discussed yet not systematically examined (
      • Strout T.D.
      • Hillen M.
      • Gutheil C.
      • Anderson E.
      • Hutchinson R.
      • Ward H.
      • Kay H.
      • Mills G.J.
      • Han P.K.J.
      Tolerance of uncertainty: A systematic review of health and healthcare-related outcomes.
      ). Informed by the sociotechnical framework, we examine factors at multiple levels, including societal (e.g., policies), technical (e.g. EHR), and cultural (e.g., staff and client race) that may influence response to changes (
      • D’Aunno T.
      • Sutton R.I.
      • Price R.H.
      Isomorphism and external support in conflicting institutional environments: A study of drug abuse treatment units.
      ). The objective of this paper is to describe program preparedness for potential changes in the addiction health services context. The research question is twofold: 1) To what extent are program leaders prepared to predict changes in the environment, and respond to changes? and 2) What are some of the program and staff factors associated with less difficulty in predicting and responding to change? This paper aims to enhance our understanding of programs reported preparedness to adapt to uncertainties in the environment, and characteristics of programs that may be emphasized in efforts to improve preparedness, and ultimately processes of care.

      2. Methods

      2.1 Data collection and procedures

      We used data from the 2014 and the 2017 National Drug Abuse Treatment System Survey (NDATSS), which is a survey of outpatient substance use disorder treatment programs (
      • D’Aunno T.
      • Pollack H.A.
      • Frimpong J.A.
      • Wutchiett D.
      Evidence-based treatment for opioid disorders: A 23-year national study of methadone dose levels.
      ,
      • D’Aunno T.
      • Pollack H.A.
      • Jiang L.
      • Metsch L.R.
      • Friedmann P.D.
      HIV testing in the nation’s opioid treatment programs, 2005–2011: The role of state regulations.
      ;
      • Friedmann P.D.
      • Lemon S.C.
      • Stein M.D.
      • D’Aunno T.A.
      Accessibility of addiction treatment: Results from a national survey of outpatient substance abuse treatment organizations.
      ). The sample included 695 and 657 programs in 2014 and in 2017, respectively. A key strength of the NDATSS is its split-panel design: each survey wave since 1988 included programs from prior waves (panel programs), and each wave also added representative samples of newer programs. The addition of new programs keeps the NDATSS representative of the changing population of US SUD treatment programs.

      2.2 Measures

      2.2.1 Dependent variable

      To make all questions consistent and in the same direction, we modified some variable definitions and recoded some responses. We examined perceived environmental uncertainty that programs might face as a result of current or potential changes in the addiction health services field, using four measures. The measures focused on how difficult program directors may find it to predict changes that will occur in their environment, or how to respond. The first measure assessed difficulty in predicting changes that would occur in the addiction health services field as a result of current or potential changes in three areas. Program directors were asked how much they agree or disagree with the following statement: “As a result of current or potential changes in the health care field, I find it hard to predict what changes will occur in …”: 1) “…regulatory requirements for our organization”, 2) “…our use of information technology such as electronic health records”, and 3) “…coordination of services for our clients.” In addition, the treatment program environment is dynamic, and these potential changes may operate concurrently, both internal and external. We therefore aimed to use a measure that captured the varying and simultaneous occurring elements of change that a program must react, and respond to. We created a composite measure of predicting change using the average of the three variables, which were based on responses to a five-point Likert scale, coded as 1 to 5 (strongly disagree – strongly agree), respectively. Hence, this first dependent variable in the composite measure was considered continuous.
      The second, third and fourth dependent variables were each measured by a single survey item. We assessed the extent to which programs find it difficult to predict the effect of change on their organization, overall, whether it has been difficult to decide how to respond to changes, and whether it is difficult to predict what changes the organization should make in response to changes in the field. In order to perform an effective and yet simple-to-interpret statistical analysis, we transformed the five-point (strongly agree – strongly disagree) into a three-level categorical variable: (disagree and strongly disagree (i.e., disagree) coded as −1; neither agree nor disagree (i.e., neutral) as 0; strongly agree and agree (i.e., agree) coded as 1). The second to fourth dependent variables were considered as ordinal categorical variables.

      2.2.2 Explanatory variables

      We included three categories of explanatory variables: program, staff, and client characteristics. Program characteristics were measured by: whether a program had a formal quality improvement plan; had electronic health record (EHR) components, which referred to an integrated electronic clinical information system that tracks patient health data, and may include functions such as encounter notes, lab order, prescriptions, etc.; if the program provides opioid treatment; ownership type (private for-profit, private not-for-profit, public); how long the program has been in operation (program age); if the program belongs to a parental organization; and, the joint commission (TJC) accreditation. Staff characteristics included: director years' experience in the field; the sources that director relied on for finding out about developments in the field of addiction health services, including participation in seminars or workshops, and informal conversations with members of other substance abuse treatment organizations. Prior studies, though scarce, have examined uncertainty tolerance in health care settings, as well as variations in treatment practices, by decision makers' characteristics. Findings from these studies have suggested variations in these outcomes by race /ethnicity of decision makers (
      • Frimpong J.A.
      • Shiu-Yee K.
      • D’Aunno T.
      The role of program directors in treatment practices: The case of methadone dose patterns in U.S. outpatient opioid agonist treatment programs.
      ;
      • Strout T.D.
      • Hillen M.
      • Gutheil C.
      • Anderson E.
      • Hutchinson R.
      • Ward H.
      • Kay H.
      • Mills G.J.
      • Han P.K.J.
      Tolerance of uncertainty: A systematic review of health and healthcare-related outcomes.
      ). We therefore assessed racial/ethnic composition of staff by the percentages of African American and Latino staff. Client characteristics included: the percentages of unemployed clients, African American clients, and Latino clients. In addition, we included the region where a program is located, to account for regional variations in policies and other guidelines.

      2.3 Data analysis

      We conducted descriptive analysis of all variables stratified by year (2014 and 2017). Chi-square tests or t-tests were conducted to determine whether dependent and independent variables were statistically different, by year. We fit linear regression and ordinal logistic regression models, using the 2014 and 2017 data, to examine associations between the dependent variables and independent variables. The variables we analyzed have low proportions of missingness, mostly under 10 %. In addition, preliminary comparative analysis suggested that most variables were not statistically significantly different between observations with and without missing values. Therefore, it seemed plausible to use the complete data by omitting records with missing independent variable in our analysis. Four regression models, one for each dependent variable, were fitted on to all the independent variables listed above in multivariable regression. Those regression models were fitted separately since the intercorrelations among the four dependent variables were low. Ordinal logistic regression was used when some dependent variables are ordinal. This approach is supported by our preliminary analysis which showed that the odds ratios (OR) for neutral relative to disagree and strong disagree were similar to those for strongly agree and agree relative to neutral for most independent variables. The statistical software STATA 17 was used to conduct all the analysis.

      3. Results

      3.1 Comparative analysis across years

      The comparative analysis of all variables by year is presented in Table 1. All four dependent variables were statistically different by year, with programs in 2017, compared with 2014, being less likely to report difficulty. Programs indicated that they found it less difficult to: predict change in their environment (mean (standard deviation): 3.3 (0.9) vs. 3.4 (0.9)), predict the effect of change on the organization (57.1 % vs. 62.4 %), decide how to respond (29.3 % vs. 36.6 %), or predict changes to make to respond (37.8 % vs. 44.3 %). For the last three dependent variables, a considerable proportion in either year, 17–25 % of programs, reported being neutral (neither agree nor disagree) on difficulty predicting or responding to change. A significantly higher proportion of programs had a quality improvement plan in 2014 than in 2017. On the other hand, there was a significant increase in the percent of programs with an electronic health record (EHR) in 2017 relative to 2014. The percent of persons who inject drugs (PWID) in treatment programs also increased over time.
      Table 1Descriptive statistics of SUD treatment programs, by year.
      2014 (N = 695)2017 (N = 657)p value
      Outcomes
       Difficulty predicting change
      p < 0.01.
      3.4 (0.9)3.3 (0.9)0.002
       Difficulty predicting effect of change on organization
      p < 0.05.
      0.044
      Strongly agree or agree418 (62.4 %)359 (57.1 %)
      Neither agree nor disagree115 (17.2 %)105 (16.7 %)
      Strongly disagree or disagree137 (20.5 %)165 (26.2 %)
       Difficulty deciding how to respond
      p < 0.05.
      0.014
      Strongly agree or agree245 (36.6 %)184 (29.3 %)
      Neither agree nor disagree159 (23.8 %)154 (24.5 %)
      Strongly disagree or disagree265 (39.6 %)290 (46.2 %)
       Difficulty predicting changes to make to respond
      p < 0.001
      0.001
      Strongly agree or agree297 (44.3 %)238 (37.8 %)
      Neither agree nor disagree157 (23.4 %)126 (20.0 %)
      Strongly disagree or disagree216 (32.2 %)265 (42.1 %)
      Program characteristics
       Quality improvement plan
      p < 0.05.
      544 (83.7 %)478 (78.9 %)0.034
       EHR
      p < 0.001
      390 (58.8 %)435 (70.5 %)0.000
       Opioid treatment programs282 (43.4 %)300 (49 %)0.051
       Type of program
      Private for-profit152 (23.3 %)162 (26.6 %)0.183
      Private not-for-profit415 (63.7 %)358 (58.7 %)0.183
      Public84 (12.9 %)90 (14.8 %)0.183
       Program age24.9 (13.8)25.6 (14.8)0.359
       Parental organization151 (23.2 %)147 (24.1 %)0.683
       TJC accreditation158 (27.4 %)141 (26.6 %)0.808
       Percent African American staff21.1 (26.0)23.6 (27.4)0.100
       Percent Latino staff10.0 (18.6)9.8 (16.6)0.801
       Region0.982
      Northeast195 (28.1 %)182 (27.7 %)
      Southeast164 (23.6 %)159 (24.2 %)
      Midwest183 (26.3 %)176 (26.8 %)
      Southwest153 (22.0 %)140 (21.3 %)
      Staff Characteristics
       Dir. years’ experience19.4 (9.9)18.9 (10.3)0.401
       Dir. seminars /workshops attendance3.6 (0.8)3.7 (0.8)0.638
       Dir. professional org. membership3.1 (1.1)3.2 (1.1)0.782
       Dir. information network3.5 (0.9)3.5 (0.9)0.989
       Race
       White461 (77.3 %)427 (73.4 %)0.443
      African American71 (11.9 %)79 (13.6 %)0.443
      Latino42 (7 %)48 (8.2 %)0.443
      Other22 (3.7 %)28 (4.8 %)0.443
      Client characteristics
       Prevalence PWID (%)
      p < 0.001
      28.1 (24.9)34.2 (27.6)0.000
       Percent unemployed clients55.6 (28.0)52.5 (28.9)0.066
       Percent African Americans19.4 (23.3)20.8 (24.7)0.304
       Percent Latino13.3 (18.6)13.4 (18.4)0.905
      low asterisk p < 0.05.
      low asterisklow asterisk p < 0.01.
      low asterisklow asterisklow asterisk p < 0.001

      3.2 Predicting change

      Table 2 shows results from the regression models. We found significant associations between program-level characteristics and SUD treatment director’s ability to predict change. Programs with EHR components had less difficult predicting changes (beta = −0.144, p < 0.05). However, programs that had been operating longer found it more difficult to predict change (beta = 0.005, p < 0.05). Programs in 2017 relative to 2014 had less difficulty predicting changes (beta = −0.056, p < 0.01). Both program and staff-level characteristics are associated with difficulty predicting effect of change on organization. Programs with a quality improvement (QI) plan had less difficulty predicting the effect of change on their organization (OR = 0.631, p < 0.05). Programs whose director had more years of experience were more likely to report difficulty predicting the effect of change (OR = 1.016, p < 0.05). On the one hand, directors who attended seminar/workshop were significantly less likely to report difficulty (OR = 0.785, p < 0.01); on the other hand, directors who seek information from their network of directors from other programs (OR = 1.189, p < 0.05) were more likely to report difficulty. Programs with a greater proportion of Latino staff (OR = 0.989, p < 0.05) were also less likely to report difficulty predicting effect of change on program. We observed that, regionally, programs in the Midwest had higher odds of difficulty in predicting the effect of change, compared with programs in Northeast (OR = 1.537, p < 0.05).
      Table 2Linear ordered logistic regression models for predicting and responding to change.
      Difficulty predicting changeDifficulty predicting effect of change on organizationDifficulty deciding how to respondDifficulty predicting changes to make to respond
      BetaCIORCIORCIORCI
      Program characteristics
       Quality improvement plan−0.019−0.172, 0.1350.631
      p < 0.05.
      0.430, 0.9241.1190.799, 1.5680.8770.626, 1.229
       EHR−0.144
      p < 0.05.
      −0.279, −0.0090.9640.698, 1.3301.1080.822, 1.4940.9270.687, 1.249
       Opioid treatment programs−0.036−0.163, 0.0910.7610.563, 1.0290.8250.622, 1.0940.9080.684, 1.206
       Type of program
      Private for-profit as reference.
      Private not-for-profit0.034−0.120, 0.1871.3190.917, 1.8981.1730.831, 1.6550.9740.692, 1.370
      Public−0.099−0.311, 0.1131.2660.763, 2.0991.634
      p < 0.05.
      1.012, 2.6381.1860.736, 1.913
       Parental organization−0.091−0.237, 0.0541.0530.744, 1.4910.9070.655, 1.2540.7560.546, 1.046
       Program age0.005
      p < 0.05.
      0.001, 0.0101.0070.996, 1.0181.011
      p < 0.05.
      1.001, 1.0211.0010.991, 1.011
       TJC accreditation−0.016−0.160, 0.1281.0640.755, 1.5001.1180.813, 1.5380.8180.594, 1.126
      Staff Characteristics
       Dir. years’ experience0.004−0.002, 0.0101.016
      p < 0.05.
      1.001, 1.0311.0040.990, 1.0181.0100.996, 1.024
       Dir. seminar/workshop attendance−0.020−0.092, 0.0520.785
      p < 0.01.
      0.658, 0.9350.838
      p < 0.05.
      0.713, 0.9840.828
      p < 0.05.
      0.703, 0.975
       Dir. Information network0.017−0.047, 0.0821.189
      p < 0.05.
      1.022, 1.3841.0140.879, 1.1691.0550.913, 1.219
       Percent African American staff−0.002−0.004, 0.0010.9970.991, 1.0040.9970.991, 1.0040.993
      p < 0.05.
      0.987, 1.000
       Percent Latino staff0.000−0.004, 0.0050.989
      p < 0.05.
      0.978, 1.0001.0010.990, 1.0110.9920.982, 1.003
      Client characteristics
       Percent unemployed clients0.0020.000, 0.0051.0020.997, 1.0081.0010.996, 1.0061.0051.000, 1.010
       Percent African American clients0.001−0.003, 0.0040.9960.988, 1.0030.9990.992, 1.0060.9990.992, 1.006
       Percent Latino clients−0.002−0.007, 0.0020.9980.987, 1.0090.988
      p < 0.05.
      0.978, 0.9990.9930.983, 1.003
      Region
      Northeast as reference.
       Southeast−0.051−0.223, 0.1200.9740.650, 1.4600.9920.678, 1.4501.1000.750, 1.613
       Midwest0.143−0.026, 0.3131.537
      p < 0.05.
      1.017, 2.3231.1510.791, 1.6740.8140.557, 1.192
       Southwest−0.048−0.227, 0.1321.0130.665, 1.5440.8920.597, 1.3320.7610.509, 1.137
      Year
      2014 as reference.
       Year−0.056
      p < 0.01.
      −0.096, −0.0160.9110.828, 1.0030.850
      p < 0.001.
      0.778, 0.9290.888
      p < 0.01.
      0.812, 0.970
      a Private for-profit as reference.
      b Northeast as reference.
      c 2014 as reference.
      low asterisk p < 0.05.
      low asterisklow asterisk p < 0.01.
      low asterisklow asterisklow asterisk p < 0.001.

      3.3 Responding to change

      Attributes of program, staff, and client characteristics were significantly associated with difficulties deciding how to respond to change. Public programs, compared with private-for-profit programs, as well as older programs reported having more difficulty deciding how to respond to current or potential changes (OR = 1.634, p < 0.05; OR = 1.011, p < 0.05, respectively). Programs whose directors attended seminar/workshop had lower odds of difficulty deciding how to respond to changes (OR = 0.838, p < 0.05). We also observed that programs with a higher percent of Latino clients were less likely to report difficulty in deciding how to respond to change (OR = 0.988, p < 0.05). Lastly, in 2017, programs reported lower odds for difficulty in deciding how to respond to changes, than in 2014 (OR = 0.850, p < 0.001). We found that programs whose directors attended seminar/workshop (OR = 0.828, p < 0.05) and programs with a higher percent of African Americans staff members (OR = 0.993, p < 0.05) had lower odds of difficulty in forecasting changes to make in order to respond. Programs were also less likely to report difficulty in predicting changes required for response, in 2017 (OR = 0.888, p < 0.01) relative to 2014.

      4. Discussion

      We examined perceived environmental uncertainty, focusing on the extent to which SUD treatment programs reported difficulty predicting and responding to changes in the addiction health services system. We focused on predicting and responding to changes in 2017, compared with 2014, as well as attributes across three main characteristics (program, staff, and clients) that may be associated with reported difficulty. We found that there were significant declines from 2014 to 2017 in the proportion of treatment programs reporting difficulty across the four dependent variables (i.e., predicting change, predicting the effect of change on the organization, deciding how to respond to change, and predicting changes to make to respond). Notwithstanding the declines, a considerable proportion of programs continued to report difficulties in 2017. For example, in 2017, a little over one half of the programs reported difficulty predicting the effect of changes on their organization, and slightly higher than one-third indicated difficulty predicting changes to make to respond. We also found that different characteristics of programs were associated with reported difficulty in predicting or responding to environmental uncertainties. Difficulty predicting change was significantly associated with program characteristics only, while program and staff characteristics were associated with difficulty predicting the effect of change on organization. Deciding how to respond was significantly associated with program, staff, and client characteristics, with only staff characteristics significantly associated with predicting changes to make to respond.
      Regarding difficulty predicting change, we found that adoption of EHR components was associated with less difficulty predicting change. Existing studies show that EHR adoption and implementation is associated with facilitating organizational responsiveness to change (
      • Callahan A.
      • Shah N.H.
      Chapter 19—Machine learning in healthcare.
      ;
      • Hamrock E.
      • Paige K.
      • Parks J.
      • Scheulen J.
      • Levin S.
      Discrete event simulation for healthcare organizations: A tool for decision making.
      ;
      • Lee S.
      • Song J.
      • Cao Q.
      Environmental uncertainty and firm performance: An empirical study with strategic alignment in the healthcare industry. ICIS 2011 proceedings.
      ;
      • Norgeot B.
      • Glicksberg B.S.
      • Butte A.J.
      A call for deep-learning healthcare.
      ). Our finding may, therefore, be related to the significant increase in programs that adopted EHR systems, and may have implemented components that facilitate documentation and tracking of processes and outcomes. The availability of these data, and perhaps related information may have equipped programs to predict changes in their operating environment. On the other hand, older programs were more likely to report more difficulty predicting change. We note that the existing literature on the effects of older program age on adaptability to change have shown mixed results, ranging from positive effects (
      • Friedmann P.D.
      • Jiang L.
      • Alexander J.A.
      Top manager effects on buprenorphine adoption in outpatient substance abuse treatment programs.
      ;
      • Roman P.M.
      • Johnson J.A.
      Adoption and implementation of new technologies in substance abuse treatment.
      ) to negative effects (
      • Lundgren L.
      • Chassler D.
      • Amodeo M.
      • D’Ippolito M.
      • Sullivan L.
      Barriers to implementation of evidence-based addiction treatment: A national study.
      ) and no effect (
      • Knudsen H.K.
      • Roman P.M.
      • Ducharme L.J.
      • Johnson J.A.
      Organizational predictors of pharmacological innovation adoption: The case of disulfiram.
      ,
      • Knudsen H.K.
      • Ducharme L.J.
      • Roman P.M.
      Early adoption of buprenorphine in substance abuse treatment centers: Data from the private and public sectors.
      ). Based on the existing literature in lines with our findings, it may be that older programs were possibly less adaptable to the changing environment. These programs may thus be less inclined to monitor changes in the environment, and consequently may not foresee uncertainties, or be positioned to decide how to respond.
      Difficulty predicting the effect of change on an organization, however, required different resources. Programs that had a formal quality improvement plan were significantly more likely to report that they found it less difficult to predict the effect of change on their organization. This finding is supported by studies on quality improvement plans in health care, which have shown positive impact on processes of care, as well as outcomes, at multiple levels (
      • Leaphart C.L.
      • Gonwa T.A.
      • Mai M.L.
      • Prendergast M.B.
      • Wadei H.M.
      • Tepas J.J.
      • Taner C.B.
      Formal quality improvement curriculum and DMAIC method results in interdisciplinary collaboration and process improvement in renal transplant patients.
      ;
      • Schouten L.M.T.
      • Hulscher M.E.J.L.
      • van Everdingen J.J.E.
      • Huijsman R.
      • Grol R.P.T.M.
      Evidence for the impact of quality improvement collaboratives: Systematic review.
      ). Several staff-level factors were also significantly associated with predicting the effect of change. Interestingly, directors with more years of experience, and larger information networks reported more difficulty predicting the effect of change on their organization, while director attendance at seminars and workshops was associated with less difficulty. These findings are in line with previous studies that report similar relationships (
      • Ducharme L.J.
      • Knudsen H.K.
      • Roman P.M.
      • Johnson J.A.
      Innovation adoption in substance abuse treatment: Exposure, trialability, and the clinical trials network.
      ;
      • Fields D.
      • Knudsen H.K.
      • Roman P.M.
      Implementation of network for the improvement of addiction treatment (NIATx) processes in substance use disorder treatment centers.
      ;
      • Friedmann P.D.
      • Jiang L.
      • Alexander J.A.
      Top manager effects on buprenorphine adoption in outpatient substance abuse treatment programs.
      ;
      • Roman P.M.
      • Johnson J.A.
      Adoption and implementation of new technologies in substance abuse treatment.
      ;
      • Savage S.A.
      • Abraham A.J.
      • Knudsen H.K.
      • Rothrauff T.C.
      • Roman P.M.
      Timing of buprenorphine adoption by privately funded substance abuse treatment programs: The role of institutional and resource-based inter-organizational linkages.
      ). We glean from these findings that external sources of information, including experts or facilitators of knowledge-based programs, may be a more useful resource for preparing programs to respond to changes. Programs with a higher proportion of Latino staff also reported less difficulty predicting the effect of change. Though we do not focus our analysis on this relationship, the association found in some studies between higher proportions of Latino staff with greater levels of cultural competence (
      • Guerrero E.G.
      Organizational characteristics that foster early adoption of cultural and linguistic competence in outpatient substance abuse treatment in the United States.
      ;
      • Guerrero E.G.
      • Campos M.
      • Urada D.
      • Yang J.C.
      Do cultural and linguistic competence matter in Latinos' completion of mandated substance abuse treatment?.
      ,
      • Guerrero E.G.
      • Khachikian T.
      • Frimpong J.A.
      • Kong Y.
      • Howard D.L.
      • Hunter S.
      Drivers of continued implementation of cultural competence in substance use disorder treatment.
      ;
      • Guerrero E.G.
      • Kim A.
      Organizational structure, leadership and readiness for change and the implementation of organizational cultural competence in addiction health services.
      ) might provide some explanation for these programs' greater confidence in their ability to predict change.
      Our results also showed that a broader range of characteristics, i.e., program, staff, and client, were associated with difficulty in deciding how to respond to change. Compared to private-for-profit programs, programs that are publicly owned were more likely to report difficulty deciding how to respond to change. Private-for-profit programs have been shown to adopt certain innovations (e.g., buprenorphine and naltrexone) at higher rates (
      • Freedman S.
      • Lin H.
      Hospital ownership type and innovation: The case of electronic medical records adoption.
      ;
      • Shields M.C.
      • Horgan C.M.
      • Ritter G.A.
      • Busch A.B.
      Use of electronic health information Technology in a National Sample of hospitals that provide specialty substance use care.
      ), while adopting other innovations (e.g. EHR) at lower rates, relative to programs of other ownership types (
      • Knudsen H.K.
      • Ducharme L.J.
      • Roman P.M.
      Early adoption of buprenorphine in substance abuse treatment centers: Data from the private and public sectors.
      ;
      • Roman P.M.
      • Abraham A.J.
      • Knudsen H.K.
      Using medication-assisted treatment for substance use disorders: Evidence of barriers and facilitators of implementation.
      ). With respect to our findings, it may be that public programs have less access to resources, i.e., information, human, financial, that promote and facilitate preparedness to act. Our finding that older age of program is associated with greater difficulty, and director’s attendance at seminars or workshops with less difficulty deciding how to respond, may operate through similar mechanisms as the other dependent variables that show significant association. It may be that due to escalation of commitment, older programs may be less likely to be aware of, acknowledge, or act in response to change. Similarly, directors who relied on seminars or workshops to find out about development in the field may have access to information and tools that better prepare them to respond, compared to directors who tap into other sources of information, i.e., reliance on other program directors. While the existing studies offer some insights, the drivers of the observed relationship with ownership type, for example, require further and targeted examination. We also found that an increasing proportion of Latino clients is associated with programs reporting less difficulty deciding how to respond. Considering our finding for the relationship between staff race/ethnicity and difficulty predicting the effect of change on organization, it would be important that future studies consider interaction effects between staff race and client race.
      Lastly, difficulty predicting changes to make to respond was only associated with staff characteristics. Programs with directors who attended seminars and workshops, as well as a greater proportion of African-American staff, tended to have decreased likelihood to report difficulty predicting changes to make to respond. Our finding that an increasing proportion of African-American staff in a program is also associated with reported increase in predicting changes to make to respond is supported by studies that have shown that higher rates of African-American and Latino healthcare staff are associated with greater patient satisfaction, as well as reduced disparities for minority patients (
      • Guerrero E.G.
      • Campos M.
      • Urada D.
      • Yang J.C.
      Do cultural and linguistic competence matter in Latinos' completion of mandated substance abuse treatment?.
      ;
      • Howard D.L.
      Culturally competent treatment of african american clients among a National Sample of outpatient substance abuse treatment units.
      ;
      • LaVeist T.A.
      • Pierre G.
      Integrating the 3Ds—Social determinants, health disparities, and health-care workforce diversity.
      ).
      There are some limitations of our findings. First, we did not link reported difficulties in predicting or responding to specific changes. It may be that programs are more prepared to predict and respond to certain types of change in the healthcare environment than others, and specificity in the question may have elicited different responses from program directors. We found, however, that a general measure of programs' reported difficulty in how they conceive of change is an important first step in improving preparedness for uncertainties. Second, our analysis examined program directors' self-reported difficulties on the outcome measures, which may not fully capture preparedness at the program level. Objective measures of responses to change in the form of changes in treatment practices would provide a more nuanced approach to understanding if and how programs respond to change. Third, our finding that older programs were more likely to report difficulty predicting change may be due to other program characteristics that are confounders for older programs and the outcomes, i.e., predicting change, and how to respond to change. Future studies should therefore seek to identify and further examine potential confounders. Lastly, we only controlled for a few program, staff, and client attributes, including client race/ethnicity, availability of EHR components, and director sources of information on advancements in the field. These factors are relevant; however, substance use disorder treatment programs are multifaceted delivery systems which may be affected by a complex set of factors and policies. This point is further illustrated by divergent findings such as the influence of program director race referenced above. Future studies should therefore examine and control for factors at multiple levels, and examine the interconnectedness across variables. This includes policies, additional structural factors, as well as processes that promote and facilitate programs' preparedness to predict changes and respond to them effectively. There is also a need for further analyses that examines internal and external environment attributes of programs, and by region, to better understand and explain our finding. Furthermore, we only considered main effects controlling for the other variables. Future studies should explore interaction effects between predictors.

      5. Conclusion

      While there are some limitations of this study, our results show that program, staff, and client levels have varying influence on predicting change and responding to change. The extent to which programs are able to predict and respond to changes in their environment also changed over time. Overall, programs have become better at assessing their environment and putting in place conditions that may facilitate making decisions about how to respond. It is promising that in 2017, compared to 2014, programs reported less difficulty in predicting change or responding to changes in the healthcare environment. Program characteristics (e.g., EHR components, age, quality improvement plan, and ownership type) are associated with predicting change, predicting effect of change on organization, deciding how to respond to change, but not predicting changes to make to respond. While increasing program age is negatively associated with predicting change, older programs are more confident in their ability to respond to change. The increasing adoption of EHR components suggest that programs will continue to make progress on these measures. To the extent older programs are positioned to predict change, i.e., availability of EHR components, they may be more likely to better decide how to respond to change. Based on these findings, we recommend that the addiction health services systems continue to support and facilitate the adoption of tools and practices associated with more favorable response uncertainty. That is, greater emphasis should be placed on the adoption of electronic health record system, and in particular equipping staff to maximize the value of this system, i.e., data driven predictions and decision making. Similarly, greater focus should be placed on programs not only planning, but implementing quality improvement initiatives, and creating opportunities for program directors to engage with and share lessons among their networks in formal settings, including workshops.
      Substance use disorder treatment programs are dynamic organizations, with various aspects of their operations influencing their ability to predict and respond to changes in their environment. Our findings are important in that it provides some indication on aspects of programs that could be centered, depending on how ready programs are to predict changes in their operating environment, and respond. Given resource constraints at multiple levels in substance use disorder treatment programs, this knowledge might help optimize the use of the available resources, and position programs to anticipate and respond to changes in a way that positively influences processes or care and patient outcomes.

      Disclosure of funding

      This work was supported by the National Institute on Minority Health and Health Disparities grant (5 R01 MD014639-01A1). The NIMHD did not have a role in study design; in the collection, analysis, and interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.

      Human participant protection statement

      This study was reviewed and approved by the Institutional Review Board of the Texas A&M University (IRB2019-0268D). One of the principal investigators, Erick G. Guerrero, has obtained consent to publish from the participants in this study (program staff members).

      CRediT authorship contribution statement

      Jemima A. Frimpong: Conceptualization, Methodology, Writing - original draft, Writing - review & editing. Erick G. Guerrero: Investigation, Funding acquisition, Writing - review & editing. Yinfei Kong: Formal analysis, Methodology, Writing - original draft, Writing - review & editing. Tenie Khachikian: Project administration, Writing - review & editing. Suojin Wang: Formal analysis, Methodology, Writing - review & editing. Thomas D'Aunno: Investigation, Data curation, Writing – review & editing. Daniel L. Howard: Investigation, Funding acquisition, Writing - review & editing.

      Declaration of competing interest

      No conflicts of interest.

      Acknowledgements

      The authors would like to thank treatment providers for their participation in this study and appreciate Mona Zahir, Research Assistant, for proofreading and preparing this paper for publication.

      References

        • Callahan A.
        • Shah N.H.
        Chapter 19—Machine learning in healthcare.
        in: Sheikh A. Cresswell K.M. Wright A. Bates D.W. Key advances in clinical informatics. Academic Press, 2017: 279-291https://doi.org/10.1016/B978-0-12-809523-2.00019-4
        • Carlo A.D.
        • Benson N.M.
        • Chu F.
        • Busch A.B.
        Association of Alternative Payment and Delivery Models with Outcomes for mental health and substance use disorders: A systematic review.
        JAMA Network Open. 2020; 3e207401https://doi.org/10.1001/jamanetworkopen.2020.7401
        • D’Aunno T.
        • Pollack H.A.
        • Frimpong J.A.
        • Wutchiett D.
        Evidence-based treatment for opioid disorders: A 23-year national study of methadone dose levels.
        Journal of Substance Abuse Treatment. 2014; 47: 245-250https://doi.org/10.1016/j.jsat.2014.06.001
        • D’Aunno T.
        • Pollack H.A.
        • Jiang L.
        • Metsch L.R.
        • Friedmann P.D.
        HIV testing in the nation’s opioid treatment programs, 2005–2011: The role of state regulations.
        Health Services Research. 2014; 49: 230-248https://doi.org/10.1111/1475-6773.12094
        • D’Aunno T.
        • Sutton R.I.
        • Price R.H.
        Isomorphism and external support in conflicting institutional environments: A study of drug abuse treatment units.
        Academy of Management Journal. 1991; 34: 636-661https://doi.org/10.2307/256409
        • Ducharme L.J.
        • Knudsen H.K.
        • Roman P.M.
        • Johnson J.A.
        Innovation adoption in substance abuse treatment: Exposure, trialability, and the clinical trials network.
        Journal of Substance Abuse Treatment. 2007; 32: 321-329https://doi.org/10.1016/j.jsat.2006.05.021
        • Duncan R.B.
        Characteristics of organizational environments and perceived environmental uncertainty.
        Administrative Science Quarterly. 1972; 17: 313-327https://doi.org/10.2307/2392145
        • Faupel S.
        • Süß S.
        The effect of transformational leadership on employees during organizational change – an empirical analysis.
        Journal of Change Management. 2019; 19: 145-166https://doi.org/10.1080/14697017.2018.1447006
        • Fields D.
        • Knudsen H.K.
        • Roman P.M.
        Implementation of network for the improvement of addiction treatment (NIATx) processes in substance use disorder treatment centers.
        The Journal of Behavioral Health Services & Research. 2016; 43: 354-365https://doi.org/10.1007/s11414-015-9466-7
        • Freedman S.
        • Lin H.
        Hospital ownership type and innovation: The case of electronic medical records adoption.
        Nonprofit and Voluntary Sector Quarterly. 2018; 47: 537-561https://doi.org/10.1177/0899764018757025
        • Friedmann P.D.
        • Jiang L.
        • Alexander J.A.
        Top manager effects on buprenorphine adoption in outpatient substance abuse treatment programs.
        The Journal of Behavioral Health Services & Research. 2010; 37: 322-337https://doi.org/10.1007/s11414-009-9169-z
        • Friedmann P.D.
        • Lemon S.C.
        • Stein M.D.
        • D’Aunno T.A.
        Accessibility of addiction treatment: Results from a national survey of outpatient substance abuse treatment organizations.
        Health Services Research. 2003; 38: 887-903https://doi.org/10.1111/1475-6773.00151
        • Frimpong J.A.
        • Jackson B.E.
        • Stewart L.M.
        • Singh K.P.
        • Rivers P.A.
        • Bae S.
        Health information technology capacity at federally qualified health centers: A mechanism for improving quality of care.
        BMC Health Services Research. 2013; 13: 35https://doi.org/10.1186/1472-6963-13-35
        • Frimpong J.A.
        • Shiu-Yee K.
        • D’Aunno T.
        The role of program directors in treatment practices: The case of methadone dose patterns in U.S. outpatient opioid agonist treatment programs.
        Health Services Research. 2017; 52: 1881-1907https://doi.org/10.1111/1475-6773.12558
        • Green C.A.
        • McCarty D.
        • Mertens J.
        • Lynch F.L.
        • Hilde A.
        • Firemark A.
        • Weisner C.M.
        • Pating D.
        • Anderson B.M.
        A qualitative study of the adoption of buprenorphine for opioid addiction treatment.
        Journal of Substance Abuse Treatment. 2014; 46: 390-401https://doi.org/10.1016/j.jsat.2013.09.002
        • Guerrero E.G.
        Organizational characteristics that foster early adoption of cultural and linguistic competence in outpatient substance abuse treatment in the United States.
        Evaluation and Program Planning. 2012; 35: 9-15https://doi.org/10.1016/j.evalprogplan.2011.06.001
        • Guerrero E.G.
        • Campos M.
        • Urada D.
        • Yang J.C.
        Do cultural and linguistic competence matter in Latinos' completion of mandated substance abuse treatment?.
        Substance Abuse Treatment, Prevention, and Policy. 2012; 7: 34https://doi.org/10.1186/1747-597X-7-34
        • Guerrero E.G.
        • Khachikian T.
        • Frimpong J.A.
        • Kong Y.
        • Howard D.L.
        • Hunter S.
        Drivers of continued implementation of cultural competence in substance use disorder treatment.
        Journal of Substance Abuse Treatment. 2019; 105: 5-11https://doi.org/10.1016/j.jsat.2019.07.009
        • Guerrero E.G.
        • Kim A.
        Organizational structure, leadership and readiness for change and the implementation of organizational cultural competence in addiction health services.
        Evaluation and Program Planning. 2013; 40: 74-81https://doi.org/10.1016/j.evalprogplan.2013.05.002
        • Hamrock E.
        • Paige K.
        • Parks J.
        • Scheulen J.
        • Levin S.
        Discrete event simulation for healthcare organizations: A tool for decision making.
        Journal of Healthcare Management. 2013; 58: 110-124
        • Han P.K.J.
        • Klein W.M.P.
        • Arora N.K.
        Varieties of uncertainty in health care: A conceptual taxonomy.
        Medical Decision Making : An International Journal of the Society for Medical Decision Making. 2011; 31: 828-838https://doi.org/10.1177/0272989X11393976
        • Howard D.L.
        Culturally competent treatment of african american clients among a National Sample of outpatient substance abuse treatment units.
        Journal of Substance Abuse Treatment. 2003; 24: 89-102https://doi.org/10.1016/S0740-5472(02)00348-3
        • Knudsen H.K.
        • Ducharme L.J.
        • Roman P.M.
        Early adoption of buprenorphine in substance abuse treatment centers: Data from the private and public sectors.
        Journal of Substance Abuse Treatment. 2006; 30: 363-373https://doi.org/10.1016/j.jsat.2006.03.013
        • Knudsen H.K.
        • Roman P.M.
        • Ducharme L.J.
        • Johnson J.A.
        Organizational predictors of pharmacological innovation adoption: The case of disulfiram.
        Journal of Drug Issues. 2005; 35: 559-573https://doi.org/10.1177/002204260503500308
        • Kreiser P.
        • Marino L.
        Analyzing the historical development of the environmental uncertainty construct.
        Management Decision. 2002; 40: 895-905https://doi.org/10.1108/00251740210441090
        • LaVeist T.A.
        • Pierre G.
        Integrating the 3Ds—Social determinants, health disparities, and health-care workforce diversity.
        Public Health Reports. 2014; 129: 9-14
        • Leaphart C.L.
        • Gonwa T.A.
        • Mai M.L.
        • Prendergast M.B.
        • Wadei H.M.
        • Tepas J.J.
        • Taner C.B.
        Formal quality improvement curriculum and DMAIC method results in interdisciplinary collaboration and process improvement in renal transplant patients.
        Journal of Surgical Research. 2012; 177: 7-13https://doi.org/10.1016/j.jss.2012.03.017
        • Lee S.
        • Song J.
        • Cao Q.
        Environmental uncertainty and firm performance: An empirical study with strategic alignment in the healthcare industry. ICIS 2011 proceedings.
        Thirty Second International Conference on Information Systems, Shanghai2011, December 6
        • Longenecker C.O.
        • Longenecker P.D.
        Why hospital improvement efforts fail: A view from the front line.
        Journal of Healthcare Management. 2014; 59: 147-157
        • Lundgren L.
        • Chassler D.
        • Amodeo M.
        • D’Ippolito M.
        • Sullivan L.
        Barriers to implementation of evidence-based addiction treatment: A national study.
        Journal of Substance Abuse Treatment. 2012; 42: 231-238https://doi.org/10.1016/j.jsat.2011.08.003
        • Lurie N.
        • Wasserman J.
        • Nelson C.D.
        Public health preparedness: Evolution or revolution?.
        Health Affairs (Project Hope). 2006; 25: 935-945https://doi.org/10.1377/hlthaff.25.4.935
        • McAlearney A.S.
        • Walker D.M.
        • Hefner J.L.
        Moving organizational culture from volume to value: A qualitative analysis of private sector accountable care organization development.
        Health Services Research. 2018; 53: 4767-4788https://doi.org/10.1111/1475-6773.13012
        • Norgeot B.
        • Glicksberg B.S.
        • Butte A.J.
        A call for deep-learning healthcare.
        Nature Medicine. 2019; 25: 14-15https://doi.org/10.1038/s41591-018-0320-3
        • Polonsky M.S.
        High-reliability organizations: The next frontier in healthcare quality and safety.
        Journal of Healthcare Management. 2019; 64: 213-221https://doi.org/10.1097/JHM-D-19-00098
        • Roman P.M.
        • Abraham A.J.
        • Knudsen H.K.
        Using medication-assisted treatment for substance use disorders: Evidence of barriers and facilitators of implementation.
        Addictive Behaviors. 2011; 36: 584-589https://doi.org/10.1016/j.addbeh.2011.01.032
        • Roman P.M.
        • Johnson J.A.
        Adoption and implementation of new technologies in substance abuse treatment.
        Journal of Substance Abuse Treatment. 2002; 22: 211-218https://doi.org/10.1016/S0740-5472(02)00241-6
        • Savage S.A.
        • Abraham A.J.
        • Knudsen H.K.
        • Rothrauff T.C.
        • Roman P.M.
        Timing of buprenorphine adoption by privately funded substance abuse treatment programs: The role of institutional and resource-based inter-organizational linkages.
        Journal of Substance Abuse Treatment. 2012; 42: 16-24https://doi.org/10.1016/j.jsat.2011.06.009
        • Schouten L.M.T.
        • Hulscher M.E.J.L.
        • van Everdingen J.J.E.
        • Huijsman R.
        • Grol R.P.T.M.
        Evidence for the impact of quality improvement collaboratives: Systematic review.
        British Medical Journal. 2008; 336: 1491-1494https://doi.org/10.1136/bmj.39570.749884.BE
        • Shields M.C.
        • Horgan C.M.
        • Ritter G.A.
        • Busch A.B.
        Use of electronic health information Technology in a National Sample of hospitals that provide specialty substance use care.
        Psychiatric Services. 2021; (appi.ps.202000816)https://doi.org/10.1176/appi.ps.202000816
        • Strout T.D.
        • Hillen M.
        • Gutheil C.
        • Anderson E.
        • Hutchinson R.
        • Ward H.
        • Kay H.
        • Mills G.J.
        • Han P.K.J.
        Tolerance of uncertainty: A systematic review of health and healthcare-related outcomes.
        Patient Education and Counseling. 2018; 101: 1518-1537https://doi.org/10.1016/j.pec.2018.03.030
        • Substance Abuse and Mental Health Services Administration a.
        • Office of the Surgeon General
        Chapter 6: Health care systems and substance use disorders.
        in: Facing addiction in America: The surgeon General’s report on alcohol, drugs, and health. US Department of Health and Human Services, 2016
        • Tai B.
        • Volkow N.D.
        Treatment for substance use disorder: Opportunities and challenges under the affordable care act.
        Social Work in Public Health. 2013; 28: 165-174https://doi.org/10.1080/19371918.2013.758975
        • Thakur R.
        • Hsu S.H.Y.
        • Fontenot G.
        Innovation in healthcare: Issues and future trends.
        Journal of Business Research. 2012; 65: 562-569https://doi.org/10.1016/j.jbusres.2011.02.022