Specifying an implementation framework for Veterans Affairs antimicrobial stewardship programmes: using a factor analysis approach

Specifying an implementation framework for Veterans Affairs antimicrobial stewardship programmes:... Abstract Objectives Inappropriate antibiotic use poses a serious threat to patient safety. Antimicrobial stewardship programmes (ASPs) may optimize antimicrobial use and improve patient outcomes, but their implementation remains an organizational challenge. Using the Promoting Action on Research Implementation in Health Services (PARiHS) framework, this study aimed to identify organizational factors that may facilitate ASP design, development and implementation. Methods Among 130 Veterans Affairs facilities that offered acute care, we classified organizational variables supporting antimicrobial stewardship activities into three PARiHS domains: evidence to encompass sources of knowledge; contexts to translate evidence into practice; and facilitation to enhance the implementation process. We conducted a series of exploratory factor analyses to identify conceptually linked factor scales. Cronbach's alphas were calculated. Variables with large uniqueness values were left as single factors. Results We identified 32 factors, including six constructs derived from factor analyses under the three PARiHS domains. In the evidence domain, four factors described guidelines and clinical pathways. The context domain was broken into three main categories: (i) receptive context (15 factors describing resources, affiliations/networks, formalized policies/practices, decision-making, receptiveness to change); (ii) team functioning (1 factor); and (iii) evaluation/feedback (5 factors). Within facilitation, two factors described facilitator roles and tasks and five captured skills and training. Conclusions We mapped survey data onto PARiHS domains to identify factors that may be adapted to facilitate ASP uptake. Our model encompasses mostly mutable factors whose relationships with performance outcomes may be explored to optimize antimicrobial use. Our framework also provides an analytical model for determining whether leveraging existing organizational processes can potentially optimize ASP performance. Introduction Appropriate use of antimicrobials is essential in ensuring optimal patient outcomes and safety. However, up to half of all antimicrobial use in the US is inappropriate, leading to increases in antimicrobial resistance (AR),1,2 adverse events, incidence of Clostridium difficile-associated colitis3 and costs.4 This problem is compounded by the dearth of new antimicrobial development that has resulted in the emergence of untreatable infections.5 Recognizing the urgency to mitigate the AR crisis, the CDC, Society for Healthcare Epidemiology of America (SHEA) and other professional societies have strongly encouraged the implementation of antimicrobial stewardship programmes (ASPs). An ASP uses a multidisciplinary approach to promote appropriate selection, dosing, route and duration of antimicrobial therapy. Despite the potential benefits of ASPs, their implementation remains a serious challenge for many healthcare systems.6,7 Research thus far has identified a wide range of barriers, such as resource constraints, lack of leadership support and participant resistance, but also characteristics specific to individual healthcare facility microsystems.6,7 Moreover, formal study of ASP implementation is in its early stages and there is substantial lack of clarity regarding what facilitates ASP uptake and effectiveness. This study aims to specify factors for constructing a framework of ASP components in order to identify corresponding strategies to facilitate ASP design, development and implementation. Methods Framework Implementation is ‘the systematic integration of research findings and other evidence-based practices into routine practice to improve the quality and effectiveness of healthcare’.8 In the context of antimicrobial stewardship (AS), specifying factors in its implementation will help develop strategies to accelerate implementation and optimize antimicrobial use.9 To guide the characterization of ASP components, we used domains from the Promoting Action on Research Implementation in Health Services (PARiHS) framework. PARiHS organizes ASP features into the following domains: evidence, context and facilitation.10–14 Evidence encompasses codified and non-codified sources of knowledge, including research evidence, clinical experience, professional craft knowledge, patient preferences and experiences, and local information. Contextual characteristics are key to ensuring a more conducive environment for programme implementation. Important characteristics may include supportive leadership, organizational structure and evaluative systems. Facilitation in this framework emphasizes the characteristics and role of the individual facilitator. The type of facilitation for programme success, and the role and skill of the facilitator that are required, is determined by the ‘state of preparedness’ of an individual or team, in terms of their acceptance and understanding of the evidence and the receptivity of their place of work or context in terms of its resources, culture and values, leadership style and evaluation activity.10,13 These domains capture the clinical implications and organizational drivers for implementing ASPs. Setting The Veterans Affairs (VA) Healthcare System provides an ideal setting to develop a framework for ASP implementation. The VA cares for military veterans and is the largest integrated healthcare system in the US, and a leader in implementing collaborative multicentre evidence-based quality improvement projects.5,15,16 In 2011, the Under Secretary for Health chartered the National VA Antimicrobial Stewardship Task Force (ASTF) to guide the national initiative to optimize antimicrobial use and enhance patient safety.17 To assess pre-existing stewardship activities, ASTF conducted a detailed nationwide survey of VA ASPs in 2012, with the assistance of the VA Healthcare Analysis and Information Group (HAIG).18 Survey The nationwide survey was developed by a Technical Advisory Group comprising physicians and clinical pharmacy specialists with expertise in administration, research and patient care. The survey was pilot-tested on a representative sample of facilities with different complexity levels (i.e. degrees of specialty services offered)19 and from diverse geographical regions [i.e. Veterans Integrated Service Networks (VISNs)]. Domains of the survey included an inventory of AS activities and the larger infectious diseases (ID) community presence at each VA facility. Questions focused on facility structure, AS policy, personnel, activities/processes, resources, barriers/acceptance, an assessment of AS needs including informatics, and outcomes such as an inventory and self-assessment of programme efficacy. The survey (see Supplementary data available at JAC Online) was fielded in November and December 2012. HAIG distributed the web-based survey to each VISN Director and Chief Medical Officer for subsequent dissemination to each facility in their network using Inquisite® survey software (Allegiance Software, Inc.). All facilities providing acute and long-term care received a request that individuals most familiar with facility AS activities complete the survey. Respondents who met the inclusion criteria included chiefs of staff, medicine, ID and pharmacy. Analysis Descriptive statistics were compiled to examine frequencies and ranges of all items in the survey.18 To identify factors in an ASP implementation framework, we employed factor analyses to simplify this complex dataset and identify underlying latent variables through observed covariance patterns.20 Guided by the PARiHS domains, we selected 57 items from the HAIG survey that reflect AS practices in VA facilities. We first examined polychoric correlations among variables in the HAIG survey. We employed independent component analyses, as many questions were categorical and binary in nature, to first determine the number of factor solutions.20 The number of factors was selected based on eigenvalues >1. Factors for the implementation model represent latent constructs, which themselves represent condensed statements among a set of variables operationally defined by their factor loadings that indicate the proportion of variance accounted for by the construct. In our analysis, we also used a uniqueness factor to determine how well the items were interpreted by the construct, retaining items within a construct that had uniqueness values <0.7, which corresponds to factor loadings >0.3, thus indicating significance and salience.20–22 Cronbach's alphas were calculated for each construct to measure internal consistency by describing the extent to which items within the factor scale are interrelated and measure the same concept or construct.23–26 Variables with uniqueness values >0.7 were left as single items in the final model. In addition, by consensus we kept certain variables as single items that had loaded on a scale if they were substantively and clinically meaningful in their representation of AS. Factor scales that had only two items were converted into indices. Along with the single items and indices, we identified PARiHS constructs in each domain with conceptually linked factor scales to specify an implementation framework for ASP. In addition, we tested criterion-related validity by assessing concurrent validity, correlating factors and having a policy specifying the establishment of ASP, which is used as a proxy variable denoting implementation.27 We performed multivariable logistic regression analysis with the Least Absolute Shrinkage and Selection Operator (LASSO) approach to identify the subset of variables across three domains (evidence, organizational context and facilitation) in the PARiHS framework that may be informative for correlating ASP implementation. The final model was determined based on the optimal penalty term using 10-fold cross-validation criteria and deviance as loss function. From the soft thresholding property of the LASSO in the generalized linear regression models, the estimated regression coefficient in the penalized regression model is prone to be biased toward zero. To mitigate these biases, we reported a more unbiased estimation of the regression coefficients from unpenalized multivariable logistic regression using the selected variables in the LASSO.28,29 Results All 130 facilities within the VA system that met the inclusion criteria completed the HAIG survey (response rate = 100%). The mean operating bed size was 81. The mean number of full-time ID providers was 1.9 across these facilities. Geographic distribution of the facilities was balanced, with 20% of facilities located in the northeast and the same in southwest, 16% each in central plains and west, and 14% each in mid-south and north central USA.18 In terms of evidence, 74% of the facilities had written clinical pathways/antimicrobial therapy guidelines available and 92% reported restrictions on the use of antibiotic agents. The context domains showed that 80% had at least one ID attending physician and 38% had an AS team. Twenty-six percent had a policy to promote substitution of oral for parenteral antibiotics; 15% reported having an antimicrobial de-escalation policy; and 43% had a process for timely review of positive blood cultures. However, 67% lacked an AS business plan. To demonstrate facilitation, 72% of the facilities reported offering educational programmes for antimicrobial use and 88% used a computerized patient records system (CPRS) to facilitate AS activities.18 Our analytic approach yielded 32 factors that were mapped to the three PARiHS domains (Table 1). The 32 factors included six scales derived from factor analyses, six indices and 20 single items. In all, 4 factors were within the evidence domain, 21 were in the context domain, and 7 described facilitation. In the evidence domain, all four factors that described guidelines and clinical pathways in ID management were kept as single items based on their clinical importance. Table 1. Factors mapped to PARiHS domains   Ka  Variable typeb  Evidence   Facility offered inpatient ID consultation    ordinal   Restrictions on the use of antibiotic agents    binary   Guideline for antimicrobial duration    binary   Written clinical pathways/guidelines for specific conditions    ordinal  Context   Resources        number of full-time ID attending physicians on site    continuous    dedicated clinical pharmacist in Emergency Department    binary    intravenous home antimicrobial infusion available    binary    perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  5  factor scale    perceived benefit of types of organizational support in achieving optimal antimicrobial use  6  factor scale   Affiliations/networks        medical and pharmacy postgraduate training programs in place  3  factor scale    participation in formal AS regional collaboratives with non-VA facilities    binary   Decision-making        ID fellows and physicians involved in antibiotic approval  2  index    ID pharmacist involved in antibiotic approval    ordinal    non-ID physician involved in antibiotic approval    ordinal   Formalization        presence/duration of formal AS policy    binary    presence of specific AS policies  3  factor scale    antibiotic stop orders in place    binary    number of antibiotic-specific order sets in place    continuous   Receptiveness to change        services receptive to AS-related interventions  4  factor scale   Team functioning        efforts and authority of AS teams  5  factor scale   Evaluation and feedback        degree of dissemination and evaluation of antimicrobial outcome data  3  index    de-escalation recommendation always or usually systematically reviewed    ordinal    feedback to providers on antimicrobial use  2  index    measurement of antimicrobial use  2  index    measure of antimicrobial expenditures    binary  Facilitation   Role/task        developed AS business plan    ordinal    presence of pharmacist or ID attending physicians on acute care ward teams  2  index   Skills and training        stewardship pharmacist had ID training    binary    educational programmes for prudent antimicrobial use    binary    resources used to ensure providers received up-to-date information on antibiotic use    continuous    number of electronic resources used to facilitate stewardship    continuous    perceived utility of ASTF information dissemination  2  index    Ka  Variable typeb  Evidence   Facility offered inpatient ID consultation    ordinal   Restrictions on the use of antibiotic agents    binary   Guideline for antimicrobial duration    binary   Written clinical pathways/guidelines for specific conditions    ordinal  Context   Resources        number of full-time ID attending physicians on site    continuous    dedicated clinical pharmacist in Emergency Department    binary    intravenous home antimicrobial infusion available    binary    perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  5  factor scale    perceived benefit of types of organizational support in achieving optimal antimicrobial use  6  factor scale   Affiliations/networks        medical and pharmacy postgraduate training programs in place  3  factor scale    participation in formal AS regional collaboratives with non-VA facilities    binary   Decision-making        ID fellows and physicians involved in antibiotic approval  2  index    ID pharmacist involved in antibiotic approval    ordinal    non-ID physician involved in antibiotic approval    ordinal   Formalization        presence/duration of formal AS policy    binary    presence of specific AS policies  3  factor scale    antibiotic stop orders in place    binary    number of antibiotic-specific order sets in place    continuous   Receptiveness to change        services receptive to AS-related interventions  4  factor scale   Team functioning        efforts and authority of AS teams  5  factor scale   Evaluation and feedback        degree of dissemination and evaluation of antimicrobial outcome data  3  index    de-escalation recommendation always or usually systematically reviewed    ordinal    feedback to providers on antimicrobial use  2  index    measurement of antimicrobial use  2  index    measure of antimicrobial expenditures    binary  Facilitation   Role/task        developed AS business plan    ordinal    presence of pharmacist or ID attending physicians on acute care ward teams  2  index   Skills and training        stewardship pharmacist had ID training    binary    educational programmes for prudent antimicrobial use    binary    resources used to ensure providers received up-to-date information on antibiotic use    continuous    number of electronic resources used to facilitate stewardship    continuous    perceived utility of ASTF information dissemination  2  index  a K is the number of items that make up the factor. b The types of variables that were derived from the factor analysis included: (i) single items where they can be a binary, ordinal or continuous variable; (ii) index made up of two items where the responses may be summed; and (iii) factor scales where these are made up of multiple items and a factor score may be generated. Table 1. Factors mapped to PARiHS domains   Ka  Variable typeb  Evidence   Facility offered inpatient ID consultation    ordinal   Restrictions on the use of antibiotic agents    binary   Guideline for antimicrobial duration    binary   Written clinical pathways/guidelines for specific conditions    ordinal  Context   Resources        number of full-time ID attending physicians on site    continuous    dedicated clinical pharmacist in Emergency Department    binary    intravenous home antimicrobial infusion available    binary    perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  5  factor scale    perceived benefit of types of organizational support in achieving optimal antimicrobial use  6  factor scale   Affiliations/networks        medical and pharmacy postgraduate training programs in place  3  factor scale    participation in formal AS regional collaboratives with non-VA facilities    binary   Decision-making        ID fellows and physicians involved in antibiotic approval  2  index    ID pharmacist involved in antibiotic approval    ordinal    non-ID physician involved in antibiotic approval    ordinal   Formalization        presence/duration of formal AS policy    binary    presence of specific AS policies  3  factor scale    antibiotic stop orders in place    binary    number of antibiotic-specific order sets in place    continuous   Receptiveness to change        services receptive to AS-related interventions  4  factor scale   Team functioning        efforts and authority of AS teams  5  factor scale   Evaluation and feedback        degree of dissemination and evaluation of antimicrobial outcome data  3  index    de-escalation recommendation always or usually systematically reviewed    ordinal    feedback to providers on antimicrobial use  2  index    measurement of antimicrobial use  2  index    measure of antimicrobial expenditures    binary  Facilitation   Role/task        developed AS business plan    ordinal    presence of pharmacist or ID attending physicians on acute care ward teams  2  index   Skills and training        stewardship pharmacist had ID training    binary    educational programmes for prudent antimicrobial use    binary    resources used to ensure providers received up-to-date information on antibiotic use    continuous    number of electronic resources used to facilitate stewardship    continuous    perceived utility of ASTF information dissemination  2  index    Ka  Variable typeb  Evidence   Facility offered inpatient ID consultation    ordinal   Restrictions on the use of antibiotic agents    binary   Guideline for antimicrobial duration    binary   Written clinical pathways/guidelines for specific conditions    ordinal  Context   Resources        number of full-time ID attending physicians on site    continuous    dedicated clinical pharmacist in Emergency Department    binary    intravenous home antimicrobial infusion available    binary    perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  5  factor scale    perceived benefit of types of organizational support in achieving optimal antimicrobial use  6  factor scale   Affiliations/networks        medical and pharmacy postgraduate training programs in place  3  factor scale    participation in formal AS regional collaboratives with non-VA facilities    binary   Decision-making        ID fellows and physicians involved in antibiotic approval  2  index    ID pharmacist involved in antibiotic approval    ordinal    non-ID physician involved in antibiotic approval    ordinal   Formalization        presence/duration of formal AS policy    binary    presence of specific AS policies  3  factor scale    antibiotic stop orders in place    binary    number of antibiotic-specific order sets in place    continuous   Receptiveness to change        services receptive to AS-related interventions  4  factor scale   Team functioning        efforts and authority of AS teams  5  factor scale   Evaluation and feedback        degree of dissemination and evaluation of antimicrobial outcome data  3  index    de-escalation recommendation always or usually systematically reviewed    ordinal    feedback to providers on antimicrobial use  2  index    measurement of antimicrobial use  2  index    measure of antimicrobial expenditures    binary  Facilitation   Role/task        developed AS business plan    ordinal    presence of pharmacist or ID attending physicians on acute care ward teams  2  index   Skills and training        stewardship pharmacist had ID training    binary    educational programmes for prudent antimicrobial use    binary    resources used to ensure providers received up-to-date information on antibiotic use    continuous    number of electronic resources used to facilitate stewardship    continuous    perceived utility of ASTF information dissemination  2  index  a K is the number of items that make up the factor. b The types of variables that were derived from the factor analysis included: (i) single items where they can be a binary, ordinal or continuous variable; (ii) index made up of two items where the responses may be summed; and (iii) factor scales where these are made up of multiple items and a factor score may be generated. The context domain was broken down further into three subdomains: receptive context; team functioning; and evaluation/feedback mechanisms. The receptive context included 15 factors that lent insights into various organizational competencies that enhanced stewardship uptake. These factors described facility resources (five factors), affiliations and networks (two factors), decision-making structures (three factors), formalized policies/protocols (four factors) and receptiveness to change (one factor) (Table 2). Table 2. Factor loadings for context domain   Factor   Survey item  1  2  3  4  5  6  Perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  iv to po conversion policy  0.878            AS SharePoint site  0.753            avoidance of double anaerobic coverage policy  0.835            intervention to improve outcomes for patients with C. difficile infection policy  0.867            AS business plan  0.823            Perceived benefit of types of organizational support in achieving optimal antimicrobial use  pharmacy support    0.763          ID physician support    0.764          administration support    0.741          provider/prescriber buy-in    0.660          information technology (IT)/data tools support    0.518          educational tools support    0.768          guideline support    0.720          Medical and pharmacy programmes in place  ID fellowship programme      0.898        emergency medicine residency programme      0.992        pharmacist residency programme      0.757        Presence of specific AS intervention  timely review of positive blood cultures by the AS team        0.717      antibiotic use de-escalation        0.506      automatic ID consultations for certain conditions        0.721      policy for intervention to limit use of non-C. difficile-directed antibiotic exposure to improve outcomes for patients with C. difficile infection.        0.698      Services receptive to AS-related interventions  medicine ICU          0.766    surgical ICU          0.714    community living centre          0.662    emergency department          0.520    surgical specialties          0.838    medical specialties          0.836    Efforts and authority of AS teams  facility has an AS team            0.977  percentage of FTE ID physicians designated for stewardship            0.984  percentage of FTE clinical pharmacists/clinical pharmacist specialists designated for stewardship            0.995  clinical pharmacist oversees day-to-day operations of AS team            0.958  ID attending physician oversees day-to-day operations of AS team            0.876  Cronbach’s alpha  0.90  0.78  0.77  0.57  0.83  0.93    Factor   Survey item  1  2  3  4  5  6  Perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  iv to po conversion policy  0.878            AS SharePoint site  0.753            avoidance of double anaerobic coverage policy  0.835            intervention to improve outcomes for patients with C. difficile infection policy  0.867            AS business plan  0.823            Perceived benefit of types of organizational support in achieving optimal antimicrobial use  pharmacy support    0.763          ID physician support    0.764          administration support    0.741          provider/prescriber buy-in    0.660          information technology (IT)/data tools support    0.518          educational tools support    0.768          guideline support    0.720          Medical and pharmacy programmes in place  ID fellowship programme      0.898        emergency medicine residency programme      0.992        pharmacist residency programme      0.757        Presence of specific AS intervention  timely review of positive blood cultures by the AS team        0.717      antibiotic use de-escalation        0.506      automatic ID consultations for certain conditions        0.721      policy for intervention to limit use of non-C. difficile-directed antibiotic exposure to improve outcomes for patients with C. difficile infection.        0.698      Services receptive to AS-related interventions  medicine ICU          0.766    surgical ICU          0.714    community living centre          0.662    emergency department          0.520    surgical specialties          0.838    medical specialties          0.836    Efforts and authority of AS teams  facility has an AS team            0.977  percentage of FTE ID physicians designated for stewardship            0.984  percentage of FTE clinical pharmacists/clinical pharmacist specialists designated for stewardship            0.995  clinical pharmacist oversees day-to-day operations of AS team            0.958  ID attending physician oversees day-to-day operations of AS team            0.876  Cronbach’s alpha  0.90  0.78  0.77  0.57  0.83  0.93  iv, intravenous; po, oral; FTE, full-time employee. Table 2. Factor loadings for context domain   Factor   Survey item  1  2  3  4  5  6  Perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  iv to po conversion policy  0.878            AS SharePoint site  0.753            avoidance of double anaerobic coverage policy  0.835            intervention to improve outcomes for patients with C. difficile infection policy  0.867            AS business plan  0.823            Perceived benefit of types of organizational support in achieving optimal antimicrobial use  pharmacy support    0.763          ID physician support    0.764          administration support    0.741          provider/prescriber buy-in    0.660          information technology (IT)/data tools support    0.518          educational tools support    0.768          guideline support    0.720          Medical and pharmacy programmes in place  ID fellowship programme      0.898        emergency medicine residency programme      0.992        pharmacist residency programme      0.757        Presence of specific AS intervention  timely review of positive blood cultures by the AS team        0.717      antibiotic use de-escalation        0.506      automatic ID consultations for certain conditions        0.721      policy for intervention to limit use of non-C. difficile-directed antibiotic exposure to improve outcomes for patients with C. difficile infection.        0.698      Services receptive to AS-related interventions  medicine ICU          0.766    surgical ICU          0.714    community living centre          0.662    emergency department          0.520    surgical specialties          0.838    medical specialties          0.836    Efforts and authority of AS teams  facility has an AS team            0.977  percentage of FTE ID physicians designated for stewardship            0.984  percentage of FTE clinical pharmacists/clinical pharmacist specialists designated for stewardship            0.995  clinical pharmacist oversees day-to-day operations of AS team            0.958  ID attending physician oversees day-to-day operations of AS team            0.876  Cronbach’s alpha  0.90  0.78  0.77  0.57  0.83  0.93    Factor   Survey item  1  2  3  4  5  6  Perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  iv to po conversion policy  0.878            AS SharePoint site  0.753            avoidance of double anaerobic coverage policy  0.835            intervention to improve outcomes for patients with C. difficile infection policy  0.867            AS business plan  0.823            Perceived benefit of types of organizational support in achieving optimal antimicrobial use  pharmacy support    0.763          ID physician support    0.764          administration support    0.741          provider/prescriber buy-in    0.660          information technology (IT)/data tools support    0.518          educational tools support    0.768          guideline support    0.720          Medical and pharmacy programmes in place  ID fellowship programme      0.898        emergency medicine residency programme      0.992        pharmacist residency programme      0.757        Presence of specific AS intervention  timely review of positive blood cultures by the AS team        0.717      antibiotic use de-escalation        0.506      automatic ID consultations for certain conditions        0.721      policy for intervention to limit use of non-C. difficile-directed antibiotic exposure to improve outcomes for patients with C. difficile infection.        0.698      Services receptive to AS-related interventions  medicine ICU          0.766    surgical ICU          0.714    community living centre          0.662    emergency department          0.520    surgical specialties          0.838    medical specialties          0.836    Efforts and authority of AS teams  facility has an AS team            0.977  percentage of FTE ID physicians designated for stewardship            0.984  percentage of FTE clinical pharmacists/clinical pharmacist specialists designated for stewardship            0.995  clinical pharmacist oversees day-to-day operations of AS team            0.958  ID attending physician oversees day-to-day operations of AS team            0.876  Cronbach’s alpha  0.90  0.78  0.77  0.57  0.83  0.93  iv, intravenous; po, oral; FTE, full-time employee. Three single items and two scales described resource-related factors. The single items indicated personnel and programme resources. The two scales examined resources derived from engagement with national VA stewardship and barriers emanating from resource constraints. Five items made up the first scale, which delineated the perceived helpfulness in engaging national VA stewardship resources to develop/augment the facility’s AS policy (alpha = 0.90). The second scale included seven items, outlining perceived benefits of various organizational support in optimizing antimicrobial use (alpha = 0.78). Two factors described the facility’s affiliations and networks: one included a three-item scale, capturing training programmes in place (alpha = 0.77) and the other a single item that reported participation in stewardship regional collaboratives. Depicting decision-making structure, three factors identified the mechanism and personnel involved for approval of antimicrobial agents. Four factors, three single items and one scale, were derived from questions on formalized AS policies and protocols. The single items documented the presence of various facility-level AS policy and standardized forms to facilitate stewardship, and the four-item scale illustrated specific stewardship interventions (alpha = 0.57). A six-item scale describing receptiveness to change encompassed clinical services deemed receptive to AS-related intervention (alpha = 0.83). In the subdomain of team functioning, a five-item scale on efforts and authority of AS teams contained questions related to the presence of an AS team and degree and duration of physician and pharmacy involvement in stewardship (alpha = 0.93). The subdomain of evaluation/feedback included five factors. Two were retained as single items, detailing systematic review of de-escalation and information sharing of antimicrobial expenditures. Three indices were created to describe measurement of and feedback to providers on their antimicrobial use. The facilitation domain included seven factors in two subdomains that illustrated roles/tasks of AS personnel and needed skills and training. Pertaining to AS roles/tasks, a single item denoted the presence of an AS business plan and an index specified the availability of pharmacists and ID attending physicians on acute care ward teams. Four single items and one index were derived from the analysis that demonstrated mechanisms in place to facilitate stewardship training and information dissemination. The index reported on the perceived utility of ASTF information dissemination. In concurrent validity testing, the logistic regression with LASSO results reported three factors that were retained as those that are associated with a policy to implement an ASP: efforts and authority of stewardship teams, systematic review of de-escalation recommendation and developing a stewardship business plan (Table 3). Table 3. Variables selected from multivariable logistic regression with LASSO that correlate with ASP implementation Factor  OR  95% CI  P valuea  Efforts and authority of AS teams  1.08  1.03–1.15  0.004  De-escalation recommendation always or usually systematically reviewed  1.14  0.82–1.59  0.437  Developed AS business plan  1.95  1.04–3.69  0.039  Factor  OR  95% CI  P valuea  Efforts and authority of AS teams  1.08  1.03–1.15  0.004  De-escalation recommendation always or usually systematically reviewed  1.14  0.82–1.59  0.437  Developed AS business plan  1.95  1.04–3.69  0.039  a Although we report P values in Table 3, LASSO is a feature (e.g. predictor/risk factor) selection approach without using P values. The LASSO approach shrinks the coefficients of those unimportant predictors to zero while retaining those that are important. In other words, a predictor has predictability on outcome if and only if its coefficient has not shrunk to zero. Table 3 reports the estimation of the regression coefficients for those selected important predictors in the LASSO even if the P value does not indicate statistical significance. Table 3. Variables selected from multivariable logistic regression with LASSO that correlate with ASP implementation Factor  OR  95% CI  P valuea  Efforts and authority of AS teams  1.08  1.03–1.15  0.004  De-escalation recommendation always or usually systematically reviewed  1.14  0.82–1.59  0.437  Developed AS business plan  1.95  1.04–3.69  0.039  Factor  OR  95% CI  P valuea  Efforts and authority of AS teams  1.08  1.03–1.15  0.004  De-escalation recommendation always or usually systematically reviewed  1.14  0.82–1.59  0.437  Developed AS business plan  1.95  1.04–3.69  0.039  a Although we report P values in Table 3, LASSO is a feature (e.g. predictor/risk factor) selection approach without using P values. The LASSO approach shrinks the coefficients of those unimportant predictors to zero while retaining those that are important. In other words, a predictor has predictability on outcome if and only if its coefficient has not shrunk to zero. Table 3 reports the estimation of the regression coefficients for those selected important predictors in the LASSO even if the P value does not indicate statistical significance. Discussion Our analysis allowed the mapping of these data onto the PARiHS framework to identify 32 factors within the evidence, context and facilitation domains (Figure 1). Selecting and applying factors appropriate for a given organization may facilitate ASP implementation, which may in turn help in optimizing antimicrobial use. These findings provided support for the conceptualization of an ASP implementation model, specifying factors that may facilitate uptake. Figure 1. View largeDownload slide ASP implementation framework. Figure 1. View largeDownload slide ASP implementation framework. The specified model encompassed mostly modifiable factors whose relationships with performance outcomes can be explored and customized to improve patient safety. Factors identified from the analyses may help determine which pre-existing organizational processes and AS-specific activities can be leveraged to optimize stewardship performance. Survey responses, when interpreted within the PARiHS framework in this fashion, can inform not only the VA’s burgeoning AS movement, but also be generalizable to non-US settings, as the VA is a centralized medical care payer and provider, similar to many care systems outside the US. Other systems or organizations aiming to implement ASPs can take the same general approach. The 32 factors identified through the factor analyses can serve as a ‘menu’ from which healthcare organizations may select, adapt and apply to design and implement ASPs. As implementation of any programme or innovation carries a certain amount of uncertainty, how well these factors may promulgate the evidence for best practice, fit into the local contexts and be amenable to facilitation can also serve as bases for evaluation to address the adaptability, trialability, compatibility and observability of the ASP.30 In assessing concurrent validity, we did identify variables that were informative via their correlation to facilities having a policy establishing ASPs, including factors related to stewardship team composition and effort and presence of a stewardship business plan. These findings support ‘grass-roots’ efforts that were observed among VA facilities to establish and formalize ASPs prior to a national policy mandate for ASP establishment at all VA facilities in 2014. In 2012, the number of facilities with an ASP implementation policy was fairly low, so having a business plan and an appropriately staffed stewardship team meant that resources were available to implement an ASP. Furthermore, in a subsequent analysis, we found that a number of the factors did favourably affect multiple categories of antimicrobial-related patient care outcomes, including presence of medical and pharmacy postgraduate training programmes, number of antibiotic-specific order sets, frequency of systematic de-escalation review, presence of pharmacists or ID attending physicians on acute care ward teams and formal ID training of the lead ASP pharmacist.31 Taken together, these analyses suggest that formalization of ASP processes, facility recognition and development of ID expertise, and development of informatics-based tools that can provide decision support and identify opportunities for antimicrobial de-escalation are critical factors in ASP development. This study has a number of strengths. The development of an ASP implementation framework is timely as the AR crisis has garnered national attention and calls for action. Moreover, our study contributes to the current literature by describing a framework consisting of measurable factors with internal consistency that examines the role of evidence, context and facilitation in the successful implementation of ASPs. To date, the application of organizational theory in the implementation of best practices has been highly variable and remains generally under-utilized.32 With growing interest in implementation research, new paradigms are needed that integrate salient organizational theories into a model that can predict actionable mechanisms facilitating implementation at the organizational level.33 Specific to ASPs, a gap exists in the current knowledge base for a comprehensive evidence-based ASP implementation framework to optimize antimicrobial use. Most current literature examines the impact of one or two factors on ASP implementation at a time.34–36 This study begins to bridge this gap by offering a full range of organizational factors in stewardship, with scale development for complex organizational processes that has the potential to link antecedents of successful change to AS activities and outcomes. In addition, using a nationwide sample, our study is one of the first to specify constructs within an implementation framework that can be both accessible to implementation researchers and ‘user-friendly’ to administrators and clinicians.33 Factors identified in our framework complement findings from qualitative studies conducted by Pakyz et al.35 and Broom et al.37,38 Their work identified themes such as face-to-face communication, interprofessional networks and collaborations as facilitators whereas those related to the lack of feedback, human resources and information technology were barriers to ASP implementation. This study has a number of limitations. First, the model needs further empirical validation to ensure that it is sufficiently comprehensive in capturing all key elements. A second iteration of the HAIG survey administered in 2015 is providing additional data to enable validation of the framework by testing associations between factors and performance on antibiotic use in confirmatory factor analyses. These data will also allow us to test the predictive value (e.g. changes in antibiotic use) and consistency of the model. Second, the model as specified addresses mutable factors but we recognize that contextual variables such as size and complexity will also contribute to implementation and performance outcomes. Third, factor analyses, in general, are used to achieve scale interpretability, reproducibility and validity. Our sample size of 130 facilities limited reproducibility but the sample is nationally representative. Fourth, we used ‘having a policy establishing an ASP’ as a proxy for a facility’s decision to implement ASP. Future studies need to measure ASP implementation using established outcomes, such as acceptance, appropriateness, adoption, cost, feasibility, fidelity, penetration and sustainability to clearly articulate the relationship between organizational factors and uptake. Finally, the HAIG survey was cross-sectional and therefore did not capture measurements that are computed with longitudinal or pre–post data, such as organizational culture. Organizational culture is a critical element in facilitating uptake of any innovative practice. Although we had a measure for perception of ASP utility, our study was unable to collect information assessing learning, risk-taking and mindfulness that have been proven to be highly correlated to individual motivation and successful organizational change.39 Analysis of additional data is currently under way and future work in this area will better address this limitation. Combating AR has long been a global health issue and responses to this crisis have ranged from policy statements from the WHO to countries mandating in-hospital ASPs.40–42 Efforts to optimize antimicrobial use require an approach that needs to be coordinated across the macro level, encompassing organizational structure and competencies, and the micro level, describing facilitative factors such as individual role, skill set and training. The framework identified in this study may facilitate decision making and strategic planning for healthcare organizations to determine resource allocation and ensure successful implementation that can be sustained to ultimately enhance patient safety. Acknowledgements We gratefully acknowledge support provided by Michael Fletcher, MPA, Sarah Youn, MPH, Catherine Loc-Carrillo, PhD, and The VA HSR&D Center for the Study of Healthcare Innovation, Implementation, and Policy, Los Angeles, CA. This work was previously presented at the VA HSR&D Research Conference (Abstract #1085). Funding This work was supported by the Veterans Health Administration Office of Health Services Research and Development (HSR&D) HX-12–018, Collaborative Research to Enhance and Advance Transformation and Excellence Initiative, CRE 12–313, ‘Cognitive Support Informatics for Antimicrobial Stewardship’ (PI: P. Glassman). Transparency declarations None to declare. Supplementary data The questionnaire appears as Supplementary data at JAC Online. References 1 Costelloe C, Metcalfe C, Lovering A et al.   Effect of antibiotic prescribing in primary care on antimicrobial resistance in individual patients: systematic review and meta-analysis. BMJ  2010; 340: c2096. Google Scholar CrossRef Search ADS PubMed  2 McGowan JE. Antimicrobial stewardship—the state of the art in 2011: focus on outcome and methods. Infect Control Hosp Epidemiol  2012; 33: 331– 7. Google Scholar CrossRef Search ADS PubMed  3 Bignardi GE. Risk factors for Clostridium difficile infection. J Hosp Infect  1998; 40: 1– 15. Google Scholar CrossRef Search ADS PubMed  4 Dellit TH, Owens RC, McGowan JEJr et al.   Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis  2007; 44: 159– 77. Google Scholar CrossRef Search ADS PubMed  5 Spellberg B, Blaser M, Guidos RJ et al.   Combating antimicrobial resistance: policy recommendations to save lives. Clin Infect Dis  2011; 52 Suppl 5: S397– 428. Google Scholar CrossRef Search ADS PubMed  6 Solberg LI, Asche SE, Pawlson LG et al.   Practice systems are associated with high-quality care for diabetes. Am J Manag Care  2008; 14: 85– 92. Google Scholar PubMed  7 Solberg LI, Crain AL, Sperl-Hillen JM et al.   Care quality and implementation of the chronic care model: a quantitative study. Ann Fam Med  2006; 4: 310– 6. Google Scholar CrossRef Search ADS PubMed  8 Grimshaw JM, Eccles MP. Is evidence-based implementation of evidence-based care possible? Med J Aust  2004; 180: S50– 1. Google Scholar PubMed  9 Grimshaw J, Eccles M, Thomas R et al.   Toward evidence-based quality improvement. Evidence (and its limitations) of the effectiveness of guideline dissemination and implementation strategies 1966–1998. J Gen Intern Med  2006; 21 Suppl 2: S14– 20. Google Scholar PubMed  10 Kitson AL, Rycroft-Malone J, Harvey G et al.   Evaluating the successful implementation of evidence into practice using the PARiHS framework: theoretical and practical challenges. Implementation Sci  2008; 3: 1. Google Scholar CrossRef Search ADS   11 Rycroft-Malone J, Seers K, Chandler J et al.   The role of evidence, context, and facilitation in an implementation trial: implications for the development of the PARIHS framework. Implementation Sci  2013; 8: 28. Google Scholar CrossRef Search ADS   12 Stetler CB, Legro MW, Rycroft-Malone J et al.   Role of “external facilitation” in implementation of research findings: a qualitative evaluation of facilitation experiences in the Veterans Health Administration. Implementation Sci  2006; 1: 23. Google Scholar CrossRef Search ADS   13 Stetler CB, Legro MW, Wallace CM et al.   The role of formative evaluation in implementation research and the QUERI experience. J Gen Intern Med  2006; 21 Suppl 2: S1– 8. Google Scholar CrossRef Search ADS PubMed  14 Stetler CB, McQueen L, Demakis J et al.   An organizational framework and strategic implementation for system-level change to enhance research-based practice: QUERI Series. Implementation Sci  2008; 3: 30. Google Scholar CrossRef Search ADS   15 Goetz MB, Hoang T, Bowman C et al.   A system-wide intervention to improve HIV testing in the Veterans Health Administration. J Gen Intern Med  2008; 23: 1200– 7. Google Scholar CrossRef Search ADS PubMed  16 Jain R, Kralovic SM, Evans ME et al.   Veterans Affairs initiative to prevent methicillin-resistant Staphylococcus aureus infections. N Engl J Med  2011; 364: 1419– 30. Google Scholar CrossRef Search ADS PubMed  17 Department of Veterans Affairs. Veterans Health Administration Directive 1031: Antimicrobial Stewardship Programs . Washington, DC, USA, 2014http://www.va.gov/vhapublications/ViewPublication.asp?pub_ID=2964. 18 Chou AF, Graber CJ, Jones M et al.   Characteristics of antimicrobial stewardship programs at Veterans Affairs hospitals: results of a nationwide survey. Infect Control Hosp Epidemiol  2016; 37: 647– 54. Google Scholar CrossRef Search ADS PubMed  19 Department of Veterans Affairs. Facility Complexity Levels. 2008. http://opes.vssc.med.va.gov/FacilityComplexityLevels/Pages/default.aspx. 20 Bartholomew DJ. Factor analysis for categorical data. J R Stat Soc  1980; 42: 293– 321. 21 Flanagan M, Ramanujam R, Sutherland J et al.   Development and validation of measures to assess prevention and control of AMR in hospitals. Med Care  2007; 45: 537– 44. Google Scholar CrossRef Search ADS PubMed  22 Kline P. An Easy Guide to Factor Analysis . New York, NY, USA: Routledge, 1994. 23 Cronbach LJ. Test reliability: its meaning and determination. Psychometrika  1947; 12: 1– 16. Google Scholar CrossRef Search ADS PubMed  24 Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika  1951; 16: 297– 334. Google Scholar CrossRef Search ADS   25 Streiner DL. Starting at the beginning: an introduction to coefficient alpha and internal consistency. J Pers Assess  2003; 80: 99– 103. Google Scholar CrossRef Search ADS PubMed  26 Tavakol M, Dennick R. Post-examination analysis of objective tests. Med Teach  2011; 33: 447– 58. Google Scholar CrossRef Search ADS PubMed  27 Carmines EG, Zeller RA. Reliability and Validity Assessment . Newbury Park, CA, USA: Sage, 1979. Google Scholar CrossRef Search ADS   28 Taylor J, Tibshirani RJ. Statistical learning and selective inference. Proc Natl Acad Sci USA  2015; 112: 7629– 34. Google Scholar CrossRef Search ADS PubMed  29 Tibshirani R. Regression shrinkage and selection via the lasso. J R Stat Soc  1996; 58: 267– 88. 30 Rogers E. Diffusion of Innovations . New York, NY, USA: Free Press, 2003. 31 Graber CJ, Jones MM, Chou AF et al.   Association of inpatient antimicrobial utilization measures with antimicrobial stewardship activities and facility characteristics of Veterans Affairs medical centers. J Hosp Med  2017; 12: 301– 9. Google Scholar CrossRef Search ADS PubMed  32 Slotnick HB, Shershneva MB. Use of theory to interpret elements of change. J Contin Educ Health Prof  2002; 22: 197– 204. Google Scholar CrossRef Search ADS PubMed  33 Yano EM. The role of organizational research in implementing evidence-based practice: QUERI Series. Implementation Sci  2008; 3: 29. Google Scholar CrossRef Search ADS   34 Magedanz L, Silliprandi EM, dos Santos RP. Impact of the pharmacist on a multidisciplinary team in an antimicrobial stewardship program: a quasi-experimental study. Int J Clin Pharm  2012; 34: 290– 4. Google Scholar CrossRef Search ADS PubMed  35 Pakyz AL, Moczygemba LR, VanderWielen LM et al.   Facilitators and barriers to implementing antimicrobial stewardship strategies: results from a qualitative study. Am J Infect Control  2014; 42: S257– 63. Google Scholar CrossRef Search ADS PubMed  36 Rimawi RH, Mazer MA, Siraj DS et al.   Impact of regular collaboration between infectious diseases and critical care practitioners on antimicrobial utilization and patient outcome. Crit Care Med  2013; 41: 2099– 107. Google Scholar CrossRef Search ADS PubMed  37 Broom A, Gibson AF, Broom J et al.   Optimizing antibiotic usage in hospitals: a qualitative study of the perspectives of hospital managers. J Hosp Infect  2016; 94: 230– 5. Google Scholar CrossRef Search ADS PubMed  38 Broom J, Broom A, Plage S et al.   Barriers to uptake of antimicrobial advice in a UK hospital: a qualitative study. J Hosp Infect  2016; 93: 418– 22. Google Scholar CrossRef Search ADS PubMed  39 Helfrich CD, Weiner BJ, McKinney MM et al.   Determinants of implementation effectiveness: adapting a framework for complex innovations. Med Care Res Rev  2007; 64: 279– 303. Google Scholar CrossRef Search ADS PubMed  40 Dumartin C, Rogues AM, Amadéo B et al.   Antibiotic usage in south-western French hospitals: trends and association with antibiotic stewardship measures. J Antimicrob Chemother  2011; 66: 1631– 7. Google Scholar CrossRef Search ADS PubMed  41 Dumartin C, Rogues AM, Amadeo B et al.   Antibiotic stewardship programmes: legal framework and structure and process indicator in Southwestern French hospitals, 2005–2008. J Hosp Infect  2011; 77: 123– 8. Google Scholar CrossRef Search ADS PubMed  42 Leung E, Weil DE, Raviglione M et al.   The WHO policy package to combat antimicrobial resistance. Bull World Health Organ  2011; 89: 390– 2. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Antimicrobial Chemotherapy Oxford University Press

Specifying an implementation framework for Veterans Affairs antimicrobial stewardship programmes: using a factor analysis approach

Loading next page...
 
/lp/ou_press/specifying-an-implementation-framework-for-veterans-affairs-b0Ww0PkDi7
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For permissions, please email: journals.permissions@oup.com.
ISSN
0305-7453
eISSN
1460-2091
D.O.I.
10.1093/jac/dky207
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives Inappropriate antibiotic use poses a serious threat to patient safety. Antimicrobial stewardship programmes (ASPs) may optimize antimicrobial use and improve patient outcomes, but their implementation remains an organizational challenge. Using the Promoting Action on Research Implementation in Health Services (PARiHS) framework, this study aimed to identify organizational factors that may facilitate ASP design, development and implementation. Methods Among 130 Veterans Affairs facilities that offered acute care, we classified organizational variables supporting antimicrobial stewardship activities into three PARiHS domains: evidence to encompass sources of knowledge; contexts to translate evidence into practice; and facilitation to enhance the implementation process. We conducted a series of exploratory factor analyses to identify conceptually linked factor scales. Cronbach's alphas were calculated. Variables with large uniqueness values were left as single factors. Results We identified 32 factors, including six constructs derived from factor analyses under the three PARiHS domains. In the evidence domain, four factors described guidelines and clinical pathways. The context domain was broken into three main categories: (i) receptive context (15 factors describing resources, affiliations/networks, formalized policies/practices, decision-making, receptiveness to change); (ii) team functioning (1 factor); and (iii) evaluation/feedback (5 factors). Within facilitation, two factors described facilitator roles and tasks and five captured skills and training. Conclusions We mapped survey data onto PARiHS domains to identify factors that may be adapted to facilitate ASP uptake. Our model encompasses mostly mutable factors whose relationships with performance outcomes may be explored to optimize antimicrobial use. Our framework also provides an analytical model for determining whether leveraging existing organizational processes can potentially optimize ASP performance. Introduction Appropriate use of antimicrobials is essential in ensuring optimal patient outcomes and safety. However, up to half of all antimicrobial use in the US is inappropriate, leading to increases in antimicrobial resistance (AR),1,2 adverse events, incidence of Clostridium difficile-associated colitis3 and costs.4 This problem is compounded by the dearth of new antimicrobial development that has resulted in the emergence of untreatable infections.5 Recognizing the urgency to mitigate the AR crisis, the CDC, Society for Healthcare Epidemiology of America (SHEA) and other professional societies have strongly encouraged the implementation of antimicrobial stewardship programmes (ASPs). An ASP uses a multidisciplinary approach to promote appropriate selection, dosing, route and duration of antimicrobial therapy. Despite the potential benefits of ASPs, their implementation remains a serious challenge for many healthcare systems.6,7 Research thus far has identified a wide range of barriers, such as resource constraints, lack of leadership support and participant resistance, but also characteristics specific to individual healthcare facility microsystems.6,7 Moreover, formal study of ASP implementation is in its early stages and there is substantial lack of clarity regarding what facilitates ASP uptake and effectiveness. This study aims to specify factors for constructing a framework of ASP components in order to identify corresponding strategies to facilitate ASP design, development and implementation. Methods Framework Implementation is ‘the systematic integration of research findings and other evidence-based practices into routine practice to improve the quality and effectiveness of healthcare’.8 In the context of antimicrobial stewardship (AS), specifying factors in its implementation will help develop strategies to accelerate implementation and optimize antimicrobial use.9 To guide the characterization of ASP components, we used domains from the Promoting Action on Research Implementation in Health Services (PARiHS) framework. PARiHS organizes ASP features into the following domains: evidence, context and facilitation.10–14 Evidence encompasses codified and non-codified sources of knowledge, including research evidence, clinical experience, professional craft knowledge, patient preferences and experiences, and local information. Contextual characteristics are key to ensuring a more conducive environment for programme implementation. Important characteristics may include supportive leadership, organizational structure and evaluative systems. Facilitation in this framework emphasizes the characteristics and role of the individual facilitator. The type of facilitation for programme success, and the role and skill of the facilitator that are required, is determined by the ‘state of preparedness’ of an individual or team, in terms of their acceptance and understanding of the evidence and the receptivity of their place of work or context in terms of its resources, culture and values, leadership style and evaluation activity.10,13 These domains capture the clinical implications and organizational drivers for implementing ASPs. Setting The Veterans Affairs (VA) Healthcare System provides an ideal setting to develop a framework for ASP implementation. The VA cares for military veterans and is the largest integrated healthcare system in the US, and a leader in implementing collaborative multicentre evidence-based quality improvement projects.5,15,16 In 2011, the Under Secretary for Health chartered the National VA Antimicrobial Stewardship Task Force (ASTF) to guide the national initiative to optimize antimicrobial use and enhance patient safety.17 To assess pre-existing stewardship activities, ASTF conducted a detailed nationwide survey of VA ASPs in 2012, with the assistance of the VA Healthcare Analysis and Information Group (HAIG).18 Survey The nationwide survey was developed by a Technical Advisory Group comprising physicians and clinical pharmacy specialists with expertise in administration, research and patient care. The survey was pilot-tested on a representative sample of facilities with different complexity levels (i.e. degrees of specialty services offered)19 and from diverse geographical regions [i.e. Veterans Integrated Service Networks (VISNs)]. Domains of the survey included an inventory of AS activities and the larger infectious diseases (ID) community presence at each VA facility. Questions focused on facility structure, AS policy, personnel, activities/processes, resources, barriers/acceptance, an assessment of AS needs including informatics, and outcomes such as an inventory and self-assessment of programme efficacy. The survey (see Supplementary data available at JAC Online) was fielded in November and December 2012. HAIG distributed the web-based survey to each VISN Director and Chief Medical Officer for subsequent dissemination to each facility in their network using Inquisite® survey software (Allegiance Software, Inc.). All facilities providing acute and long-term care received a request that individuals most familiar with facility AS activities complete the survey. Respondents who met the inclusion criteria included chiefs of staff, medicine, ID and pharmacy. Analysis Descriptive statistics were compiled to examine frequencies and ranges of all items in the survey.18 To identify factors in an ASP implementation framework, we employed factor analyses to simplify this complex dataset and identify underlying latent variables through observed covariance patterns.20 Guided by the PARiHS domains, we selected 57 items from the HAIG survey that reflect AS practices in VA facilities. We first examined polychoric correlations among variables in the HAIG survey. We employed independent component analyses, as many questions were categorical and binary in nature, to first determine the number of factor solutions.20 The number of factors was selected based on eigenvalues >1. Factors for the implementation model represent latent constructs, which themselves represent condensed statements among a set of variables operationally defined by their factor loadings that indicate the proportion of variance accounted for by the construct. In our analysis, we also used a uniqueness factor to determine how well the items were interpreted by the construct, retaining items within a construct that had uniqueness values <0.7, which corresponds to factor loadings >0.3, thus indicating significance and salience.20–22 Cronbach's alphas were calculated for each construct to measure internal consistency by describing the extent to which items within the factor scale are interrelated and measure the same concept or construct.23–26 Variables with uniqueness values >0.7 were left as single items in the final model. In addition, by consensus we kept certain variables as single items that had loaded on a scale if they were substantively and clinically meaningful in their representation of AS. Factor scales that had only two items were converted into indices. Along with the single items and indices, we identified PARiHS constructs in each domain with conceptually linked factor scales to specify an implementation framework for ASP. In addition, we tested criterion-related validity by assessing concurrent validity, correlating factors and having a policy specifying the establishment of ASP, which is used as a proxy variable denoting implementation.27 We performed multivariable logistic regression analysis with the Least Absolute Shrinkage and Selection Operator (LASSO) approach to identify the subset of variables across three domains (evidence, organizational context and facilitation) in the PARiHS framework that may be informative for correlating ASP implementation. The final model was determined based on the optimal penalty term using 10-fold cross-validation criteria and deviance as loss function. From the soft thresholding property of the LASSO in the generalized linear regression models, the estimated regression coefficient in the penalized regression model is prone to be biased toward zero. To mitigate these biases, we reported a more unbiased estimation of the regression coefficients from unpenalized multivariable logistic regression using the selected variables in the LASSO.28,29 Results All 130 facilities within the VA system that met the inclusion criteria completed the HAIG survey (response rate = 100%). The mean operating bed size was 81. The mean number of full-time ID providers was 1.9 across these facilities. Geographic distribution of the facilities was balanced, with 20% of facilities located in the northeast and the same in southwest, 16% each in central plains and west, and 14% each in mid-south and north central USA.18 In terms of evidence, 74% of the facilities had written clinical pathways/antimicrobial therapy guidelines available and 92% reported restrictions on the use of antibiotic agents. The context domains showed that 80% had at least one ID attending physician and 38% had an AS team. Twenty-six percent had a policy to promote substitution of oral for parenteral antibiotics; 15% reported having an antimicrobial de-escalation policy; and 43% had a process for timely review of positive blood cultures. However, 67% lacked an AS business plan. To demonstrate facilitation, 72% of the facilities reported offering educational programmes for antimicrobial use and 88% used a computerized patient records system (CPRS) to facilitate AS activities.18 Our analytic approach yielded 32 factors that were mapped to the three PARiHS domains (Table 1). The 32 factors included six scales derived from factor analyses, six indices and 20 single items. In all, 4 factors were within the evidence domain, 21 were in the context domain, and 7 described facilitation. In the evidence domain, all four factors that described guidelines and clinical pathways in ID management were kept as single items based on their clinical importance. Table 1. Factors mapped to PARiHS domains   Ka  Variable typeb  Evidence   Facility offered inpatient ID consultation    ordinal   Restrictions on the use of antibiotic agents    binary   Guideline for antimicrobial duration    binary   Written clinical pathways/guidelines for specific conditions    ordinal  Context   Resources        number of full-time ID attending physicians on site    continuous    dedicated clinical pharmacist in Emergency Department    binary    intravenous home antimicrobial infusion available    binary    perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  5  factor scale    perceived benefit of types of organizational support in achieving optimal antimicrobial use  6  factor scale   Affiliations/networks        medical and pharmacy postgraduate training programs in place  3  factor scale    participation in formal AS regional collaboratives with non-VA facilities    binary   Decision-making        ID fellows and physicians involved in antibiotic approval  2  index    ID pharmacist involved in antibiotic approval    ordinal    non-ID physician involved in antibiotic approval    ordinal   Formalization        presence/duration of formal AS policy    binary    presence of specific AS policies  3  factor scale    antibiotic stop orders in place    binary    number of antibiotic-specific order sets in place    continuous   Receptiveness to change        services receptive to AS-related interventions  4  factor scale   Team functioning        efforts and authority of AS teams  5  factor scale   Evaluation and feedback        degree of dissemination and evaluation of antimicrobial outcome data  3  index    de-escalation recommendation always or usually systematically reviewed    ordinal    feedback to providers on antimicrobial use  2  index    measurement of antimicrobial use  2  index    measure of antimicrobial expenditures    binary  Facilitation   Role/task        developed AS business plan    ordinal    presence of pharmacist or ID attending physicians on acute care ward teams  2  index   Skills and training        stewardship pharmacist had ID training    binary    educational programmes for prudent antimicrobial use    binary    resources used to ensure providers received up-to-date information on antibiotic use    continuous    number of electronic resources used to facilitate stewardship    continuous    perceived utility of ASTF information dissemination  2  index    Ka  Variable typeb  Evidence   Facility offered inpatient ID consultation    ordinal   Restrictions on the use of antibiotic agents    binary   Guideline for antimicrobial duration    binary   Written clinical pathways/guidelines for specific conditions    ordinal  Context   Resources        number of full-time ID attending physicians on site    continuous    dedicated clinical pharmacist in Emergency Department    binary    intravenous home antimicrobial infusion available    binary    perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  5  factor scale    perceived benefit of types of organizational support in achieving optimal antimicrobial use  6  factor scale   Affiliations/networks        medical and pharmacy postgraduate training programs in place  3  factor scale    participation in formal AS regional collaboratives with non-VA facilities    binary   Decision-making        ID fellows and physicians involved in antibiotic approval  2  index    ID pharmacist involved in antibiotic approval    ordinal    non-ID physician involved in antibiotic approval    ordinal   Formalization        presence/duration of formal AS policy    binary    presence of specific AS policies  3  factor scale    antibiotic stop orders in place    binary    number of antibiotic-specific order sets in place    continuous   Receptiveness to change        services receptive to AS-related interventions  4  factor scale   Team functioning        efforts and authority of AS teams  5  factor scale   Evaluation and feedback        degree of dissemination and evaluation of antimicrobial outcome data  3  index    de-escalation recommendation always or usually systematically reviewed    ordinal    feedback to providers on antimicrobial use  2  index    measurement of antimicrobial use  2  index    measure of antimicrobial expenditures    binary  Facilitation   Role/task        developed AS business plan    ordinal    presence of pharmacist or ID attending physicians on acute care ward teams  2  index   Skills and training        stewardship pharmacist had ID training    binary    educational programmes for prudent antimicrobial use    binary    resources used to ensure providers received up-to-date information on antibiotic use    continuous    number of electronic resources used to facilitate stewardship    continuous    perceived utility of ASTF information dissemination  2  index  a K is the number of items that make up the factor. b The types of variables that were derived from the factor analysis included: (i) single items where they can be a binary, ordinal or continuous variable; (ii) index made up of two items where the responses may be summed; and (iii) factor scales where these are made up of multiple items and a factor score may be generated. Table 1. Factors mapped to PARiHS domains   Ka  Variable typeb  Evidence   Facility offered inpatient ID consultation    ordinal   Restrictions on the use of antibiotic agents    binary   Guideline for antimicrobial duration    binary   Written clinical pathways/guidelines for specific conditions    ordinal  Context   Resources        number of full-time ID attending physicians on site    continuous    dedicated clinical pharmacist in Emergency Department    binary    intravenous home antimicrobial infusion available    binary    perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  5  factor scale    perceived benefit of types of organizational support in achieving optimal antimicrobial use  6  factor scale   Affiliations/networks        medical and pharmacy postgraduate training programs in place  3  factor scale    participation in formal AS regional collaboratives with non-VA facilities    binary   Decision-making        ID fellows and physicians involved in antibiotic approval  2  index    ID pharmacist involved in antibiotic approval    ordinal    non-ID physician involved in antibiotic approval    ordinal   Formalization        presence/duration of formal AS policy    binary    presence of specific AS policies  3  factor scale    antibiotic stop orders in place    binary    number of antibiotic-specific order sets in place    continuous   Receptiveness to change        services receptive to AS-related interventions  4  factor scale   Team functioning        efforts and authority of AS teams  5  factor scale   Evaluation and feedback        degree of dissemination and evaluation of antimicrobial outcome data  3  index    de-escalation recommendation always or usually systematically reviewed    ordinal    feedback to providers on antimicrobial use  2  index    measurement of antimicrobial use  2  index    measure of antimicrobial expenditures    binary  Facilitation   Role/task        developed AS business plan    ordinal    presence of pharmacist or ID attending physicians on acute care ward teams  2  index   Skills and training        stewardship pharmacist had ID training    binary    educational programmes for prudent antimicrobial use    binary    resources used to ensure providers received up-to-date information on antibiotic use    continuous    number of electronic resources used to facilitate stewardship    continuous    perceived utility of ASTF information dissemination  2  index    Ka  Variable typeb  Evidence   Facility offered inpatient ID consultation    ordinal   Restrictions on the use of antibiotic agents    binary   Guideline for antimicrobial duration    binary   Written clinical pathways/guidelines for specific conditions    ordinal  Context   Resources        number of full-time ID attending physicians on site    continuous    dedicated clinical pharmacist in Emergency Department    binary    intravenous home antimicrobial infusion available    binary    perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  5  factor scale    perceived benefit of types of organizational support in achieving optimal antimicrobial use  6  factor scale   Affiliations/networks        medical and pharmacy postgraduate training programs in place  3  factor scale    participation in formal AS regional collaboratives with non-VA facilities    binary   Decision-making        ID fellows and physicians involved in antibiotic approval  2  index    ID pharmacist involved in antibiotic approval    ordinal    non-ID physician involved in antibiotic approval    ordinal   Formalization        presence/duration of formal AS policy    binary    presence of specific AS policies  3  factor scale    antibiotic stop orders in place    binary    number of antibiotic-specific order sets in place    continuous   Receptiveness to change        services receptive to AS-related interventions  4  factor scale   Team functioning        efforts and authority of AS teams  5  factor scale   Evaluation and feedback        degree of dissemination and evaluation of antimicrobial outcome data  3  index    de-escalation recommendation always or usually systematically reviewed    ordinal    feedback to providers on antimicrobial use  2  index    measurement of antimicrobial use  2  index    measure of antimicrobial expenditures    binary  Facilitation   Role/task        developed AS business plan    ordinal    presence of pharmacist or ID attending physicians on acute care ward teams  2  index   Skills and training        stewardship pharmacist had ID training    binary    educational programmes for prudent antimicrobial use    binary    resources used to ensure providers received up-to-date information on antibiotic use    continuous    number of electronic resources used to facilitate stewardship    continuous    perceived utility of ASTF information dissemination  2  index  a K is the number of items that make up the factor. b The types of variables that were derived from the factor analysis included: (i) single items where they can be a binary, ordinal or continuous variable; (ii) index made up of two items where the responses may be summed; and (iii) factor scales where these are made up of multiple items and a factor score may be generated. The context domain was broken down further into three subdomains: receptive context; team functioning; and evaluation/feedback mechanisms. The receptive context included 15 factors that lent insights into various organizational competencies that enhanced stewardship uptake. These factors described facility resources (five factors), affiliations and networks (two factors), decision-making structures (three factors), formalized policies/protocols (four factors) and receptiveness to change (one factor) (Table 2). Table 2. Factor loadings for context domain   Factor   Survey item  1  2  3  4  5  6  Perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  iv to po conversion policy  0.878            AS SharePoint site  0.753            avoidance of double anaerobic coverage policy  0.835            intervention to improve outcomes for patients with C. difficile infection policy  0.867            AS business plan  0.823            Perceived benefit of types of organizational support in achieving optimal antimicrobial use  pharmacy support    0.763          ID physician support    0.764          administration support    0.741          provider/prescriber buy-in    0.660          information technology (IT)/data tools support    0.518          educational tools support    0.768          guideline support    0.720          Medical and pharmacy programmes in place  ID fellowship programme      0.898        emergency medicine residency programme      0.992        pharmacist residency programme      0.757        Presence of specific AS intervention  timely review of positive blood cultures by the AS team        0.717      antibiotic use de-escalation        0.506      automatic ID consultations for certain conditions        0.721      policy for intervention to limit use of non-C. difficile-directed antibiotic exposure to improve outcomes for patients with C. difficile infection.        0.698      Services receptive to AS-related interventions  medicine ICU          0.766    surgical ICU          0.714    community living centre          0.662    emergency department          0.520    surgical specialties          0.838    medical specialties          0.836    Efforts and authority of AS teams  facility has an AS team            0.977  percentage of FTE ID physicians designated for stewardship            0.984  percentage of FTE clinical pharmacists/clinical pharmacist specialists designated for stewardship            0.995  clinical pharmacist oversees day-to-day operations of AS team            0.958  ID attending physician oversees day-to-day operations of AS team            0.876  Cronbach’s alpha  0.90  0.78  0.77  0.57  0.83  0.93    Factor   Survey item  1  2  3  4  5  6  Perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  iv to po conversion policy  0.878            AS SharePoint site  0.753            avoidance of double anaerobic coverage policy  0.835            intervention to improve outcomes for patients with C. difficile infection policy  0.867            AS business plan  0.823            Perceived benefit of types of organizational support in achieving optimal antimicrobial use  pharmacy support    0.763          ID physician support    0.764          administration support    0.741          provider/prescriber buy-in    0.660          information technology (IT)/data tools support    0.518          educational tools support    0.768          guideline support    0.720          Medical and pharmacy programmes in place  ID fellowship programme      0.898        emergency medicine residency programme      0.992        pharmacist residency programme      0.757        Presence of specific AS intervention  timely review of positive blood cultures by the AS team        0.717      antibiotic use de-escalation        0.506      automatic ID consultations for certain conditions        0.721      policy for intervention to limit use of non-C. difficile-directed antibiotic exposure to improve outcomes for patients with C. difficile infection.        0.698      Services receptive to AS-related interventions  medicine ICU          0.766    surgical ICU          0.714    community living centre          0.662    emergency department          0.520    surgical specialties          0.838    medical specialties          0.836    Efforts and authority of AS teams  facility has an AS team            0.977  percentage of FTE ID physicians designated for stewardship            0.984  percentage of FTE clinical pharmacists/clinical pharmacist specialists designated for stewardship            0.995  clinical pharmacist oversees day-to-day operations of AS team            0.958  ID attending physician oversees day-to-day operations of AS team            0.876  Cronbach’s alpha  0.90  0.78  0.77  0.57  0.83  0.93  iv, intravenous; po, oral; FTE, full-time employee. Table 2. Factor loadings for context domain   Factor   Survey item  1  2  3  4  5  6  Perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  iv to po conversion policy  0.878            AS SharePoint site  0.753            avoidance of double anaerobic coverage policy  0.835            intervention to improve outcomes for patients with C. difficile infection policy  0.867            AS business plan  0.823            Perceived benefit of types of organizational support in achieving optimal antimicrobial use  pharmacy support    0.763          ID physician support    0.764          administration support    0.741          provider/prescriber buy-in    0.660          information technology (IT)/data tools support    0.518          educational tools support    0.768          guideline support    0.720          Medical and pharmacy programmes in place  ID fellowship programme      0.898        emergency medicine residency programme      0.992        pharmacist residency programme      0.757        Presence of specific AS intervention  timely review of positive blood cultures by the AS team        0.717      antibiotic use de-escalation        0.506      automatic ID consultations for certain conditions        0.721      policy for intervention to limit use of non-C. difficile-directed antibiotic exposure to improve outcomes for patients with C. difficile infection.        0.698      Services receptive to AS-related interventions  medicine ICU          0.766    surgical ICU          0.714    community living centre          0.662    emergency department          0.520    surgical specialties          0.838    medical specialties          0.836    Efforts and authority of AS teams  facility has an AS team            0.977  percentage of FTE ID physicians designated for stewardship            0.984  percentage of FTE clinical pharmacists/clinical pharmacist specialists designated for stewardship            0.995  clinical pharmacist oversees day-to-day operations of AS team            0.958  ID attending physician oversees day-to-day operations of AS team            0.876  Cronbach’s alpha  0.90  0.78  0.77  0.57  0.83  0.93    Factor   Survey item  1  2  3  4  5  6  Perceived helpfulness in engaging national VA stewardship resources to develop/augment facility’s AS policy  iv to po conversion policy  0.878            AS SharePoint site  0.753            avoidance of double anaerobic coverage policy  0.835            intervention to improve outcomes for patients with C. difficile infection policy  0.867            AS business plan  0.823            Perceived benefit of types of organizational support in achieving optimal antimicrobial use  pharmacy support    0.763          ID physician support    0.764          administration support    0.741          provider/prescriber buy-in    0.660          information technology (IT)/data tools support    0.518          educational tools support    0.768          guideline support    0.720          Medical and pharmacy programmes in place  ID fellowship programme      0.898        emergency medicine residency programme      0.992        pharmacist residency programme      0.757        Presence of specific AS intervention  timely review of positive blood cultures by the AS team        0.717      antibiotic use de-escalation        0.506      automatic ID consultations for certain conditions        0.721      policy for intervention to limit use of non-C. difficile-directed antibiotic exposure to improve outcomes for patients with C. difficile infection.        0.698      Services receptive to AS-related interventions  medicine ICU          0.766    surgical ICU          0.714    community living centre          0.662    emergency department          0.520    surgical specialties          0.838    medical specialties          0.836    Efforts and authority of AS teams  facility has an AS team            0.977  percentage of FTE ID physicians designated for stewardship            0.984  percentage of FTE clinical pharmacists/clinical pharmacist specialists designated for stewardship            0.995  clinical pharmacist oversees day-to-day operations of AS team            0.958  ID attending physician oversees day-to-day operations of AS team            0.876  Cronbach’s alpha  0.90  0.78  0.77  0.57  0.83  0.93  iv, intravenous; po, oral; FTE, full-time employee. Three single items and two scales described resource-related factors. The single items indicated personnel and programme resources. The two scales examined resources derived from engagement with national VA stewardship and barriers emanating from resource constraints. Five items made up the first scale, which delineated the perceived helpfulness in engaging national VA stewardship resources to develop/augment the facility’s AS policy (alpha = 0.90). The second scale included seven items, outlining perceived benefits of various organizational support in optimizing antimicrobial use (alpha = 0.78). Two factors described the facility’s affiliations and networks: one included a three-item scale, capturing training programmes in place (alpha = 0.77) and the other a single item that reported participation in stewardship regional collaboratives. Depicting decision-making structure, three factors identified the mechanism and personnel involved for approval of antimicrobial agents. Four factors, three single items and one scale, were derived from questions on formalized AS policies and protocols. The single items documented the presence of various facility-level AS policy and standardized forms to facilitate stewardship, and the four-item scale illustrated specific stewardship interventions (alpha = 0.57). A six-item scale describing receptiveness to change encompassed clinical services deemed receptive to AS-related intervention (alpha = 0.83). In the subdomain of team functioning, a five-item scale on efforts and authority of AS teams contained questions related to the presence of an AS team and degree and duration of physician and pharmacy involvement in stewardship (alpha = 0.93). The subdomain of evaluation/feedback included five factors. Two were retained as single items, detailing systematic review of de-escalation and information sharing of antimicrobial expenditures. Three indices were created to describe measurement of and feedback to providers on their antimicrobial use. The facilitation domain included seven factors in two subdomains that illustrated roles/tasks of AS personnel and needed skills and training. Pertaining to AS roles/tasks, a single item denoted the presence of an AS business plan and an index specified the availability of pharmacists and ID attending physicians on acute care ward teams. Four single items and one index were derived from the analysis that demonstrated mechanisms in place to facilitate stewardship training and information dissemination. The index reported on the perceived utility of ASTF information dissemination. In concurrent validity testing, the logistic regression with LASSO results reported three factors that were retained as those that are associated with a policy to implement an ASP: efforts and authority of stewardship teams, systematic review of de-escalation recommendation and developing a stewardship business plan (Table 3). Table 3. Variables selected from multivariable logistic regression with LASSO that correlate with ASP implementation Factor  OR  95% CI  P valuea  Efforts and authority of AS teams  1.08  1.03–1.15  0.004  De-escalation recommendation always or usually systematically reviewed  1.14  0.82–1.59  0.437  Developed AS business plan  1.95  1.04–3.69  0.039  Factor  OR  95% CI  P valuea  Efforts and authority of AS teams  1.08  1.03–1.15  0.004  De-escalation recommendation always or usually systematically reviewed  1.14  0.82–1.59  0.437  Developed AS business plan  1.95  1.04–3.69  0.039  a Although we report P values in Table 3, LASSO is a feature (e.g. predictor/risk factor) selection approach without using P values. The LASSO approach shrinks the coefficients of those unimportant predictors to zero while retaining those that are important. In other words, a predictor has predictability on outcome if and only if its coefficient has not shrunk to zero. Table 3 reports the estimation of the regression coefficients for those selected important predictors in the LASSO even if the P value does not indicate statistical significance. Table 3. Variables selected from multivariable logistic regression with LASSO that correlate with ASP implementation Factor  OR  95% CI  P valuea  Efforts and authority of AS teams  1.08  1.03–1.15  0.004  De-escalation recommendation always or usually systematically reviewed  1.14  0.82–1.59  0.437  Developed AS business plan  1.95  1.04–3.69  0.039  Factor  OR  95% CI  P valuea  Efforts and authority of AS teams  1.08  1.03–1.15  0.004  De-escalation recommendation always or usually systematically reviewed  1.14  0.82–1.59  0.437  Developed AS business plan  1.95  1.04–3.69  0.039  a Although we report P values in Table 3, LASSO is a feature (e.g. predictor/risk factor) selection approach without using P values. The LASSO approach shrinks the coefficients of those unimportant predictors to zero while retaining those that are important. In other words, a predictor has predictability on outcome if and only if its coefficient has not shrunk to zero. Table 3 reports the estimation of the regression coefficients for those selected important predictors in the LASSO even if the P value does not indicate statistical significance. Discussion Our analysis allowed the mapping of these data onto the PARiHS framework to identify 32 factors within the evidence, context and facilitation domains (Figure 1). Selecting and applying factors appropriate for a given organization may facilitate ASP implementation, which may in turn help in optimizing antimicrobial use. These findings provided support for the conceptualization of an ASP implementation model, specifying factors that may facilitate uptake. Figure 1. View largeDownload slide ASP implementation framework. Figure 1. View largeDownload slide ASP implementation framework. The specified model encompassed mostly modifiable factors whose relationships with performance outcomes can be explored and customized to improve patient safety. Factors identified from the analyses may help determine which pre-existing organizational processes and AS-specific activities can be leveraged to optimize stewardship performance. Survey responses, when interpreted within the PARiHS framework in this fashion, can inform not only the VA’s burgeoning AS movement, but also be generalizable to non-US settings, as the VA is a centralized medical care payer and provider, similar to many care systems outside the US. Other systems or organizations aiming to implement ASPs can take the same general approach. The 32 factors identified through the factor analyses can serve as a ‘menu’ from which healthcare organizations may select, adapt and apply to design and implement ASPs. As implementation of any programme or innovation carries a certain amount of uncertainty, how well these factors may promulgate the evidence for best practice, fit into the local contexts and be amenable to facilitation can also serve as bases for evaluation to address the adaptability, trialability, compatibility and observability of the ASP.30 In assessing concurrent validity, we did identify variables that were informative via their correlation to facilities having a policy establishing ASPs, including factors related to stewardship team composition and effort and presence of a stewardship business plan. These findings support ‘grass-roots’ efforts that were observed among VA facilities to establish and formalize ASPs prior to a national policy mandate for ASP establishment at all VA facilities in 2014. In 2012, the number of facilities with an ASP implementation policy was fairly low, so having a business plan and an appropriately staffed stewardship team meant that resources were available to implement an ASP. Furthermore, in a subsequent analysis, we found that a number of the factors did favourably affect multiple categories of antimicrobial-related patient care outcomes, including presence of medical and pharmacy postgraduate training programmes, number of antibiotic-specific order sets, frequency of systematic de-escalation review, presence of pharmacists or ID attending physicians on acute care ward teams and formal ID training of the lead ASP pharmacist.31 Taken together, these analyses suggest that formalization of ASP processes, facility recognition and development of ID expertise, and development of informatics-based tools that can provide decision support and identify opportunities for antimicrobial de-escalation are critical factors in ASP development. This study has a number of strengths. The development of an ASP implementation framework is timely as the AR crisis has garnered national attention and calls for action. Moreover, our study contributes to the current literature by describing a framework consisting of measurable factors with internal consistency that examines the role of evidence, context and facilitation in the successful implementation of ASPs. To date, the application of organizational theory in the implementation of best practices has been highly variable and remains generally under-utilized.32 With growing interest in implementation research, new paradigms are needed that integrate salient organizational theories into a model that can predict actionable mechanisms facilitating implementation at the organizational level.33 Specific to ASPs, a gap exists in the current knowledge base for a comprehensive evidence-based ASP implementation framework to optimize antimicrobial use. Most current literature examines the impact of one or two factors on ASP implementation at a time.34–36 This study begins to bridge this gap by offering a full range of organizational factors in stewardship, with scale development for complex organizational processes that has the potential to link antecedents of successful change to AS activities and outcomes. In addition, using a nationwide sample, our study is one of the first to specify constructs within an implementation framework that can be both accessible to implementation researchers and ‘user-friendly’ to administrators and clinicians.33 Factors identified in our framework complement findings from qualitative studies conducted by Pakyz et al.35 and Broom et al.37,38 Their work identified themes such as face-to-face communication, interprofessional networks and collaborations as facilitators whereas those related to the lack of feedback, human resources and information technology were barriers to ASP implementation. This study has a number of limitations. First, the model needs further empirical validation to ensure that it is sufficiently comprehensive in capturing all key elements. A second iteration of the HAIG survey administered in 2015 is providing additional data to enable validation of the framework by testing associations between factors and performance on antibiotic use in confirmatory factor analyses. These data will also allow us to test the predictive value (e.g. changes in antibiotic use) and consistency of the model. Second, the model as specified addresses mutable factors but we recognize that contextual variables such as size and complexity will also contribute to implementation and performance outcomes. Third, factor analyses, in general, are used to achieve scale interpretability, reproducibility and validity. Our sample size of 130 facilities limited reproducibility but the sample is nationally representative. Fourth, we used ‘having a policy establishing an ASP’ as a proxy for a facility’s decision to implement ASP. Future studies need to measure ASP implementation using established outcomes, such as acceptance, appropriateness, adoption, cost, feasibility, fidelity, penetration and sustainability to clearly articulate the relationship between organizational factors and uptake. Finally, the HAIG survey was cross-sectional and therefore did not capture measurements that are computed with longitudinal or pre–post data, such as organizational culture. Organizational culture is a critical element in facilitating uptake of any innovative practice. Although we had a measure for perception of ASP utility, our study was unable to collect information assessing learning, risk-taking and mindfulness that have been proven to be highly correlated to individual motivation and successful organizational change.39 Analysis of additional data is currently under way and future work in this area will better address this limitation. Combating AR has long been a global health issue and responses to this crisis have ranged from policy statements from the WHO to countries mandating in-hospital ASPs.40–42 Efforts to optimize antimicrobial use require an approach that needs to be coordinated across the macro level, encompassing organizational structure and competencies, and the micro level, describing facilitative factors such as individual role, skill set and training. The framework identified in this study may facilitate decision making and strategic planning for healthcare organizations to determine resource allocation and ensure successful implementation that can be sustained to ultimately enhance patient safety. Acknowledgements We gratefully acknowledge support provided by Michael Fletcher, MPA, Sarah Youn, MPH, Catherine Loc-Carrillo, PhD, and The VA HSR&D Center for the Study of Healthcare Innovation, Implementation, and Policy, Los Angeles, CA. This work was previously presented at the VA HSR&D Research Conference (Abstract #1085). Funding This work was supported by the Veterans Health Administration Office of Health Services Research and Development (HSR&D) HX-12–018, Collaborative Research to Enhance and Advance Transformation and Excellence Initiative, CRE 12–313, ‘Cognitive Support Informatics for Antimicrobial Stewardship’ (PI: P. Glassman). Transparency declarations None to declare. Supplementary data The questionnaire appears as Supplementary data at JAC Online. References 1 Costelloe C, Metcalfe C, Lovering A et al.   Effect of antibiotic prescribing in primary care on antimicrobial resistance in individual patients: systematic review and meta-analysis. BMJ  2010; 340: c2096. Google Scholar CrossRef Search ADS PubMed  2 McGowan JE. Antimicrobial stewardship—the state of the art in 2011: focus on outcome and methods. Infect Control Hosp Epidemiol  2012; 33: 331– 7. Google Scholar CrossRef Search ADS PubMed  3 Bignardi GE. Risk factors for Clostridium difficile infection. J Hosp Infect  1998; 40: 1– 15. Google Scholar CrossRef Search ADS PubMed  4 Dellit TH, Owens RC, McGowan JEJr et al.   Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis  2007; 44: 159– 77. Google Scholar CrossRef Search ADS PubMed  5 Spellberg B, Blaser M, Guidos RJ et al.   Combating antimicrobial resistance: policy recommendations to save lives. Clin Infect Dis  2011; 52 Suppl 5: S397– 428. Google Scholar CrossRef Search ADS PubMed  6 Solberg LI, Asche SE, Pawlson LG et al.   Practice systems are associated with high-quality care for diabetes. Am J Manag Care  2008; 14: 85– 92. Google Scholar PubMed  7 Solberg LI, Crain AL, Sperl-Hillen JM et al.   Care quality and implementation of the chronic care model: a quantitative study. Ann Fam Med  2006; 4: 310– 6. Google Scholar CrossRef Search ADS PubMed  8 Grimshaw JM, Eccles MP. Is evidence-based implementation of evidence-based care possible? Med J Aust  2004; 180: S50– 1. Google Scholar PubMed  9 Grimshaw J, Eccles M, Thomas R et al.   Toward evidence-based quality improvement. Evidence (and its limitations) of the effectiveness of guideline dissemination and implementation strategies 1966–1998. J Gen Intern Med  2006; 21 Suppl 2: S14– 20. Google Scholar PubMed  10 Kitson AL, Rycroft-Malone J, Harvey G et al.   Evaluating the successful implementation of evidence into practice using the PARiHS framework: theoretical and practical challenges. Implementation Sci  2008; 3: 1. Google Scholar CrossRef Search ADS   11 Rycroft-Malone J, Seers K, Chandler J et al.   The role of evidence, context, and facilitation in an implementation trial: implications for the development of the PARIHS framework. Implementation Sci  2013; 8: 28. Google Scholar CrossRef Search ADS   12 Stetler CB, Legro MW, Rycroft-Malone J et al.   Role of “external facilitation” in implementation of research findings: a qualitative evaluation of facilitation experiences in the Veterans Health Administration. Implementation Sci  2006; 1: 23. Google Scholar CrossRef Search ADS   13 Stetler CB, Legro MW, Wallace CM et al.   The role of formative evaluation in implementation research and the QUERI experience. J Gen Intern Med  2006; 21 Suppl 2: S1– 8. Google Scholar CrossRef Search ADS PubMed  14 Stetler CB, McQueen L, Demakis J et al.   An organizational framework and strategic implementation for system-level change to enhance research-based practice: QUERI Series. Implementation Sci  2008; 3: 30. Google Scholar CrossRef Search ADS   15 Goetz MB, Hoang T, Bowman C et al.   A system-wide intervention to improve HIV testing in the Veterans Health Administration. J Gen Intern Med  2008; 23: 1200– 7. Google Scholar CrossRef Search ADS PubMed  16 Jain R, Kralovic SM, Evans ME et al.   Veterans Affairs initiative to prevent methicillin-resistant Staphylococcus aureus infections. N Engl J Med  2011; 364: 1419– 30. Google Scholar CrossRef Search ADS PubMed  17 Department of Veterans Affairs. Veterans Health Administration Directive 1031: Antimicrobial Stewardship Programs . Washington, DC, USA, 2014http://www.va.gov/vhapublications/ViewPublication.asp?pub_ID=2964. 18 Chou AF, Graber CJ, Jones M et al.   Characteristics of antimicrobial stewardship programs at Veterans Affairs hospitals: results of a nationwide survey. Infect Control Hosp Epidemiol  2016; 37: 647– 54. Google Scholar CrossRef Search ADS PubMed  19 Department of Veterans Affairs. Facility Complexity Levels. 2008. http://opes.vssc.med.va.gov/FacilityComplexityLevels/Pages/default.aspx. 20 Bartholomew DJ. Factor analysis for categorical data. J R Stat Soc  1980; 42: 293– 321. 21 Flanagan M, Ramanujam R, Sutherland J et al.   Development and validation of measures to assess prevention and control of AMR in hospitals. Med Care  2007; 45: 537– 44. Google Scholar CrossRef Search ADS PubMed  22 Kline P. An Easy Guide to Factor Analysis . New York, NY, USA: Routledge, 1994. 23 Cronbach LJ. Test reliability: its meaning and determination. Psychometrika  1947; 12: 1– 16. Google Scholar CrossRef Search ADS PubMed  24 Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika  1951; 16: 297– 334. Google Scholar CrossRef Search ADS   25 Streiner DL. Starting at the beginning: an introduction to coefficient alpha and internal consistency. J Pers Assess  2003; 80: 99– 103. Google Scholar CrossRef Search ADS PubMed  26 Tavakol M, Dennick R. Post-examination analysis of objective tests. Med Teach  2011; 33: 447– 58. Google Scholar CrossRef Search ADS PubMed  27 Carmines EG, Zeller RA. Reliability and Validity Assessment . Newbury Park, CA, USA: Sage, 1979. Google Scholar CrossRef Search ADS   28 Taylor J, Tibshirani RJ. Statistical learning and selective inference. Proc Natl Acad Sci USA  2015; 112: 7629– 34. Google Scholar CrossRef Search ADS PubMed  29 Tibshirani R. Regression shrinkage and selection via the lasso. J R Stat Soc  1996; 58: 267– 88. 30 Rogers E. Diffusion of Innovations . New York, NY, USA: Free Press, 2003. 31 Graber CJ, Jones MM, Chou AF et al.   Association of inpatient antimicrobial utilization measures with antimicrobial stewardship activities and facility characteristics of Veterans Affairs medical centers. J Hosp Med  2017; 12: 301– 9. Google Scholar CrossRef Search ADS PubMed  32 Slotnick HB, Shershneva MB. Use of theory to interpret elements of change. J Contin Educ Health Prof  2002; 22: 197– 204. Google Scholar CrossRef Search ADS PubMed  33 Yano EM. The role of organizational research in implementing evidence-based practice: QUERI Series. Implementation Sci  2008; 3: 29. Google Scholar CrossRef Search ADS   34 Magedanz L, Silliprandi EM, dos Santos RP. Impact of the pharmacist on a multidisciplinary team in an antimicrobial stewardship program: a quasi-experimental study. Int J Clin Pharm  2012; 34: 290– 4. Google Scholar CrossRef Search ADS PubMed  35 Pakyz AL, Moczygemba LR, VanderWielen LM et al.   Facilitators and barriers to implementing antimicrobial stewardship strategies: results from a qualitative study. Am J Infect Control  2014; 42: S257– 63. Google Scholar CrossRef Search ADS PubMed  36 Rimawi RH, Mazer MA, Siraj DS et al.   Impact of regular collaboration between infectious diseases and critical care practitioners on antimicrobial utilization and patient outcome. Crit Care Med  2013; 41: 2099– 107. Google Scholar CrossRef Search ADS PubMed  37 Broom A, Gibson AF, Broom J et al.   Optimizing antibiotic usage in hospitals: a qualitative study of the perspectives of hospital managers. J Hosp Infect  2016; 94: 230– 5. Google Scholar CrossRef Search ADS PubMed  38 Broom J, Broom A, Plage S et al.   Barriers to uptake of antimicrobial advice in a UK hospital: a qualitative study. J Hosp Infect  2016; 93: 418– 22. Google Scholar CrossRef Search ADS PubMed  39 Helfrich CD, Weiner BJ, McKinney MM et al.   Determinants of implementation effectiveness: adapting a framework for complex innovations. Med Care Res Rev  2007; 64: 279– 303. Google Scholar CrossRef Search ADS PubMed  40 Dumartin C, Rogues AM, Amadéo B et al.   Antibiotic usage in south-western French hospitals: trends and association with antibiotic stewardship measures. J Antimicrob Chemother  2011; 66: 1631– 7. Google Scholar CrossRef Search ADS PubMed  41 Dumartin C, Rogues AM, Amadeo B et al.   Antibiotic stewardship programmes: legal framework and structure and process indicator in Southwestern French hospitals, 2005–2008. J Hosp Infect  2011; 77: 123– 8. Google Scholar CrossRef Search ADS PubMed  42 Leung E, Weil DE, Raviglione M et al.   The WHO policy package to combat antimicrobial resistance. Bull World Health Organ  2011; 89: 390– 2. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of Antimicrobial ChemotherapyOxford University Press

Published: Jun 4, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off