JAMDA 16 (2015) 648e653
JAMDA journal homepage: www.jamda.com
Original Study
Process Evaluation of a Quality Improvement Project to Decrease Hospital Readmissions From Skilled Nursing Facilities Thomas P. Meehan Sr. MD, MPH a, b, *, Daniel J. Qazi BS a, Thomas J. Van Hoof MD, EdD a, c, d, Shih-Yieh Ho PhD, MPH a, Sheila Eckenrode MA, RN a, Ann Spenard MSN, RN a, Michelle Pandolfi MSW, MBA a, Florence Johnson MSN, MHA, RN a, Deborah Quetti MBA, RN a a
Qualidigm, Wethersfield, CT Department of Health Sciences, Frank H. Netter MD School of Medicine, North Haven, CT University of Connecticut School of Nursing, Storrs, CT d Department of Community Medicine and Healthcare, University of Connecticut School of Medicine, Farmington, CT b c
a b s t r a c t Keywords: Skilled nursing facility quality improvement process evaluation
Objective: To describe and evaluate the impact of quality improvement (QI) support provided to skilled nursing facilities (SNFs) by a Quality Improvement Organization (QIO). Design: Retrospective, mixed-method, process evaluation of a QI project intended to decrease preventable hospital readmissions from SNFs. Setting: Five SNFs in Connecticut. Participants: SNF Administrators, Directors of Nursing, Assistant Directors of Nursing, Admissions Coordinators, Registered Nurses, Certified Nursing Assistants, Receptionists, QIO Quality Improvement Consultant. Intervention: QIO staff provided training and technical assistance to SNF administrative and clinical staff to establish or enhance QI infrastructure and implement an established set of QI tools [Interventions to Reduce Acute Care Transfers (INTERACT) tools]. Measurements: Baseline SNF demographic, staffing, and hospital readmission data; baseline and followup SNF QI structure (QI Committee), processes (general and use of INTERACT tools), and outcome (30-day all-cause hospital readmission rates); details of QIO-provided training and technical assistance; QIOperceived barriers to quality improvement; SNF leadership-perceived barriers, accomplishments, and suggestions for improvement of QIO support. Results: Success occurred in establishing QI Committees and targeting preventable hospital readmissions, as well as implementing INTERACT tools in all SNFs; however, hospital readmission rates decreased in only 2 facilities. QIO staff and SNF leaders noted the ongoing challenge of engaging already busy SNF staff and leadership in QI activities. SNF leaders reported that they appreciated the training and technical assistance that their institutions received, although most noted that additional support was needed to bring about improvement in readmission rates. Conclusion: This process evaluation documented mixed clinical results but successfully identified opportunities to improve recruitment of and provision of technical support to participating SNFs. Recommendations are offered for others who wish to conduct similar projects. Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Qualidigm’s work was funded by the Centers for Medicare and Medicaid Services through its 10th Scope of Work Quality Improvement Organization contract (Contract number: HHSM-500-2011-CT10C). This material was prepared by Qualidigm, the 10th Scope of Work Medicare Quality Improvement Organization for Connecticut, under contract with the Centers for Medicare and Medicaid Services (CMS), an agency of the U.S. Department of http://dx.doi.org/10.1016/j.jamda.2015.02.015 1525-8610/Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Health and Human Services. The contents do not necessarily reflect CMS policy. Pub # CT-6133-2014132. The authors declare no conflicts of interest. * Address correspondence to Thomas P. Meehan Sr., MD, MPH, Qualidigm, 1290 Silas Deane Highway, Suite 4A, Wethersfield, CT 06109-4337. E-mail address: [email protected] (T.P. Meehan).
T.P. Meehan Sr. et al. / JAMDA 16 (2015) 648e653
The Centers for Medicare and Medicaid Services (CMS) has implemented a variety of strategies to improve quality of care and to reduce costs in all settings of care. An important focus has been the reduction of preventable hospital readmissions. In 2011, approximately 19% of all hospitalized Medicare beneficiaries were readmitted to a hospital within 30 days of discharge.1 That same year, approximately 25% of Medicare beneficiaries discharged from the hospital to a skilled nursing facility (SNF) were re-hospitalized at least once during the year, costing Medicare $14.3 billion.1 Some CMS interventions to decrease hospital readmissions have included public reporting of hospital readmission rates and financial penalties for hospitals whose readmission rates are higher than expected based on their patient mix. Public reporting of skilled nursing facility readmission rates is scheduled to begin in 2017 and financial penalties for skilled nursing facilities with higher than expected readmission rates will be levied in 2018.2 In addition to public reporting and reduced payments, CMS has provided quality improvement (QI) training and technical assistance to healthcare providers through its network of Quality Improvement Organizations (QIOs).3 CMS referred to this component of the QIOs’ 10th Statement of Work contract as Care Transitions. The Connecticut QIO addressed care transitions by providing training and technical assistance to communities of healthcare and social service providers within defined hospital service areas.4,5 One component of this QIO’s work was a pilot project involving 5 SNFs in a single-hospital community. QIO staff worked with personnel from the SNFs to reduce hospital readmission rates by providing general QI training and targeted assistance in the use of a publically available QI program called Interventions to Reduce Acute Care Transfers (INTERACT). The program focuses on improving the identification, evaluation, and management of acute changes in the condition of SNF residents and provides a variety of tools to assist with these goals. Several studies have been published describing the INTERACT program and its impact on hospitalization rates.6e8 These studies document substantial decreases in hospitalization rates among the participating institutions and provide a general description of support provided to SNFs to implement the program. However, details are lacking on how an external organization, such as a QIO, can recruit SNFs to participate in a QI project involving INTERACT and can facilitate use of the INTERACT tools to decrease hospital readmission rates. We conducted a process evaluation of this pilot QI project with 3 goals: (1) to describe our steps in recruiting SNFs, training SNF staff, and assisting SNF staff with adoption of the INTERACT tools; (2) to determine changes in quality improvement structure and processes, and a specific outcome (30-day all-cause hospital readmission) in the participating SNFs; and (3) to document barriers, successes, and suggested enhancements to our training and technical assistance. Methods Recruitment To identify potential participants, a QIO analyst identified all SNFs within the target community using addresses listed on Medicare’s Nursing Home Compare website9 and calculated 30-day all-cause hospital readmission rates for each facility based on claims data for fee-for-service Medicare beneficiaries for the first 6 months of 2013. Participation criteria included having 40 beds and a 30-day readmission rate of 10% among fee-for-service Medicare beneficiaries during this period. Next, a QIO staff member with an established track record of QI consulting in Connecticut SNFs mailed a recruitment letter to each SNF Administrator or Director of Nursing. This letter was cosigned by the hospital representative who was the designated leader of the community effort to decrease preventable hospital
649
readmissions. The letter emphasized the national focus on reduction of preventable hospital readmissions and Medicare’s strategies of public reporting, financial penalties, and technical assistance from QIOs. Next, 2 other QIO staff members who also had histories of QI consulting to CT SNFs placed follow-up phone calls. These calls were to request a face-to-face meeting with the Administrator, the Director of Nursing, and the Medical Director to discuss the project, answer questions, and obtain leadership commitment to participate. These meetings occurred between July, 1 and November 30, 2013. Quality Improvement Training and Technical Assistance Two QI consultants from the QIO provided individualized training and technical assistance to designated personnel from the participating SNFs from the date of their commitment to the project through June 30, 2014. From July through December, this training focused on use of the INTERACT Hospitalization Tracking Tool. During this phase, the first QI consultant made monthly visits to the SNFs to demonstrate use of the tool, check accuracy of data entry, and review data output. From January through June, the second QI consultant interacted with staff from each SNF to identify opportunities for improvement and to facilitate implementation of the INTERACT tools. All participating SNFs were asked to implement the INTERACT Hospitalization Tracking Tool, the Quality Improvement Tool for Review of Acute Care Transfers, the SituationeBackgroundeAssessmenteRecommendation (SBAR) Communication Form and Progress Note, and the Stop and Watch Tool; additional INTERACT tools were recommended by the QI Consultant based on the specific needs of each institution. Both QI Consultants supplemented their face-to-face visits with phone calls and e-mails, as needed. Data Collection and Analysis During the July through December 2013, training of designated SNF personnel on the INTERACT Hospitalization Tracking tool, the first QI Consultant collected baseline SNF-reported data on resident hospitalizations. From January 1 through June 30 2014, SNF staff collected these data independently and provided them to the QI Consultant on a monthly basis. In December 2013, the second QI Consultant conducted a SNF Needs Assessment that captured administrative information (number of beds, types of units, occupancy rates, for-profit vs nonprofit status, membership in a corporate chain); electronic health record implementation status; tenure of SNF leaders (Administrator, Director of Nursing Services, Medical Director); numbers of physicians, Advanced Practice Registered Nurses; and staffing ratios (minutes per resident day) for registered nurses, certified nursing assistants, and physical therapists; existence of a QI Committee; and use of INTERACT tools. QIO staff recorded quantitative and qualitative data on their interactions with SNF personnel in an electronic database referred to as the Contact Log. They captured dates of interactions, QIO and SNF personnel involved, method of interaction (face-to-face, phone, email, or fax), purpose of interaction (project overview, participation agreement, needs assessment, data collection training, data retrieval, data accuracy assessment, performance data feedback, root cause analysis of resident transfers, staff education, and technical assistance related to INTERACT tool use), duration of contact in minutes, barriers to use of INTERACT tools (competing staff responsibilities, staff availability, staff turnover, clinician resistance to change, inadequate staff communication, technology challenges), use of INTERACT tools prior to the project (Hospitalization Tracking Tool, QI Tool for Review of Acute Care Transfers, Stop and Watch Early Warning Tool, SBAR Communication Form and Progress Note, Facility Capabilities List, Care Paths), and ad hoc notes. In addition to the Contact Log, the QIO team produced internal team meeting minutes that summarized
650
T.P. Meehan Sr. et al. / JAMDA 16 (2015) 648e653
team discussions and problem solving. Finally, the 2 QI Consultants jointly conducted semistructured exit interviews with leaders from each SNF at the conclusion of the project. These interviews elicited SNF leaders’ perspectives on staff use of INTERACT data collection and intervention tools, perceived barriers and benefits to their use, and suggestions for improvement of the QIO training and technical assistance. Both QIO staff members kept independent notes on the interviews; afterwards, they compared notes and reconciled differences in a final written document. Analysis began with calculations of univariate frequencies and simple descriptive statistics for all quantitative data. This included staffing ratios from Nursing Home Compare postings of November 2013, SNF Needs Assessment variables, Contact Log entries, Hospitalization Tracking Tool entries, and quantitative data from the exit interviews with SNF leadership. Changes in QI structure (presence of a QI Committee) and general QI processes (tracking of resident hospitalizations, root cause analysis of resident hospitalizations, and sharing of SNF capabilities with local hospitals) and use of INTERACT tools were assessed by comparing data from the needs assessments and the exit interviews. SNF-specific 30-day all-cause hospital readmission rates were calculated from the Hospitalization Tracking Tool data using the date of admission to the SNF as day zero and using rolling-quarters to compensate for small denominators. Results were graphed in run charts. Statistical significance testing was not done because the numbers were small. Analyses were conducted using Excel, version 2010 (Microsoft, Redmond, WA) and SAS, v 9.2 32-bit (SAS Institute, Inc, Cary, NC). Internal team meeting minutes and SNF leadership exit interview notes were reviewed to validate Contact Log entries and to identify SNF leader suggestions for improvement of QIO-provided training and technical assistance.
Results Recruitment QIO staff mailed recruitment letters to the 15 SNFs within the community that met the selection criteria. Then, QIO personnel made 1e3 phone calls to the Directors of Nursing Services and the Administrators of each SNF to schedule on-site recruitment visits. Leaders at 7 facilities agreed to meet with QIO staff to discuss the project. SNF leaders who were present at the meetings most commonly included the Administrator and the Director of Nursing Services, although others, including the Assistant Director of Nursing Services, the Director of Admissions, the Staff Development Director, and the Infection Preventionist attended at some sites. The Medical Director was not present at any of these initial meetings. Leaders from all 7 SNFs that met with QIO staff agreed to participate in the project. However, both the Administrator and the Director of Nursing Services at 2 SNFs left their positions subsequent to these meetings and these SNFs withdrew from the project. Leaders from the 8 SNFs that did not meet with QIO staff either did not answer the QIO’s initial phone calls or declined to participate, citing lack of time and competing priorities as their reasons. Many of these SNF leaders expressed concern when QIO personnel mentioned that the project would involve use of an electronic hospitalization tracking tool, stating that they were unsure of their staff’s technologic capabilities. In addition, these SNF leaders indicated that they did not see a compelling reason to participate in the project. Table 1 summarizes demographic characteristics and pre-intervention all-cause 30-day readmission rates of the 5 SNFs that participated in the project. Of note, SNF A had an Administrator that had been in place much longer than the other participating facilities; SNF E had much higher staffing ratios; and SNF C had shorter tenures for all of its leadership staff, and
Table 1 Administrative Characteristics of Participating SNFs Characteristic
Beds per unit: Long-term Short-term Dementia care Percent occupancy For-profit Corporate chain Electronic health record Leadership tenure in years: Administrator Director of nursing services Medical director Staffing: Admitting physicians (including medical director) Advanced practice registered nurses Nurse minutes per resident day Certified nursing assistant minutes per resident day Physical therapist minutes per resident day *Pre-intervention 30-day all cause readmission rates (%)
SNF A
B
C
D
E
62 14 95 Yes No Yes
102 30 90 Yes No No
85 30 35 90 Yes Yes No
120 30 95 Yes Yes Yes
53 12 93 Yes No No
30 5 10
5 6 20
JAMDA journal homepage: www.jamda.com
Original Study
Process Evaluation of a Quality Improvement Project to Decrease Hospital Readmissions From Skilled Nursing Facilities Thomas P. Meehan Sr. MD, MPH a, b, *, Daniel J. Qazi BS a, Thomas J. Van Hoof MD, EdD a, c, d, Shih-Yieh Ho PhD, MPH a, Sheila Eckenrode MA, RN a, Ann Spenard MSN, RN a, Michelle Pandolfi MSW, MBA a, Florence Johnson MSN, MHA, RN a, Deborah Quetti MBA, RN a a
Qualidigm, Wethersfield, CT Department of Health Sciences, Frank H. Netter MD School of Medicine, North Haven, CT University of Connecticut School of Nursing, Storrs, CT d Department of Community Medicine and Healthcare, University of Connecticut School of Medicine, Farmington, CT b c
a b s t r a c t Keywords: Skilled nursing facility quality improvement process evaluation
Objective: To describe and evaluate the impact of quality improvement (QI) support provided to skilled nursing facilities (SNFs) by a Quality Improvement Organization (QIO). Design: Retrospective, mixed-method, process evaluation of a QI project intended to decrease preventable hospital readmissions from SNFs. Setting: Five SNFs in Connecticut. Participants: SNF Administrators, Directors of Nursing, Assistant Directors of Nursing, Admissions Coordinators, Registered Nurses, Certified Nursing Assistants, Receptionists, QIO Quality Improvement Consultant. Intervention: QIO staff provided training and technical assistance to SNF administrative and clinical staff to establish or enhance QI infrastructure and implement an established set of QI tools [Interventions to Reduce Acute Care Transfers (INTERACT) tools]. Measurements: Baseline SNF demographic, staffing, and hospital readmission data; baseline and followup SNF QI structure (QI Committee), processes (general and use of INTERACT tools), and outcome (30-day all-cause hospital readmission rates); details of QIO-provided training and technical assistance; QIOperceived barriers to quality improvement; SNF leadership-perceived barriers, accomplishments, and suggestions for improvement of QIO support. Results: Success occurred in establishing QI Committees and targeting preventable hospital readmissions, as well as implementing INTERACT tools in all SNFs; however, hospital readmission rates decreased in only 2 facilities. QIO staff and SNF leaders noted the ongoing challenge of engaging already busy SNF staff and leadership in QI activities. SNF leaders reported that they appreciated the training and technical assistance that their institutions received, although most noted that additional support was needed to bring about improvement in readmission rates. Conclusion: This process evaluation documented mixed clinical results but successfully identified opportunities to improve recruitment of and provision of technical support to participating SNFs. Recommendations are offered for others who wish to conduct similar projects. Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Qualidigm’s work was funded by the Centers for Medicare and Medicaid Services through its 10th Scope of Work Quality Improvement Organization contract (Contract number: HHSM-500-2011-CT10C). This material was prepared by Qualidigm, the 10th Scope of Work Medicare Quality Improvement Organization for Connecticut, under contract with the Centers for Medicare and Medicaid Services (CMS), an agency of the U.S. Department of http://dx.doi.org/10.1016/j.jamda.2015.02.015 1525-8610/Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Health and Human Services. The contents do not necessarily reflect CMS policy. Pub # CT-6133-2014132. The authors declare no conflicts of interest. * Address correspondence to Thomas P. Meehan Sr., MD, MPH, Qualidigm, 1290 Silas Deane Highway, Suite 4A, Wethersfield, CT 06109-4337. E-mail address: [email protected] (T.P. Meehan).
T.P. Meehan Sr. et al. / JAMDA 16 (2015) 648e653
The Centers for Medicare and Medicaid Services (CMS) has implemented a variety of strategies to improve quality of care and to reduce costs in all settings of care. An important focus has been the reduction of preventable hospital readmissions. In 2011, approximately 19% of all hospitalized Medicare beneficiaries were readmitted to a hospital within 30 days of discharge.1 That same year, approximately 25% of Medicare beneficiaries discharged from the hospital to a skilled nursing facility (SNF) were re-hospitalized at least once during the year, costing Medicare $14.3 billion.1 Some CMS interventions to decrease hospital readmissions have included public reporting of hospital readmission rates and financial penalties for hospitals whose readmission rates are higher than expected based on their patient mix. Public reporting of skilled nursing facility readmission rates is scheduled to begin in 2017 and financial penalties for skilled nursing facilities with higher than expected readmission rates will be levied in 2018.2 In addition to public reporting and reduced payments, CMS has provided quality improvement (QI) training and technical assistance to healthcare providers through its network of Quality Improvement Organizations (QIOs).3 CMS referred to this component of the QIOs’ 10th Statement of Work contract as Care Transitions. The Connecticut QIO addressed care transitions by providing training and technical assistance to communities of healthcare and social service providers within defined hospital service areas.4,5 One component of this QIO’s work was a pilot project involving 5 SNFs in a single-hospital community. QIO staff worked with personnel from the SNFs to reduce hospital readmission rates by providing general QI training and targeted assistance in the use of a publically available QI program called Interventions to Reduce Acute Care Transfers (INTERACT). The program focuses on improving the identification, evaluation, and management of acute changes in the condition of SNF residents and provides a variety of tools to assist with these goals. Several studies have been published describing the INTERACT program and its impact on hospitalization rates.6e8 These studies document substantial decreases in hospitalization rates among the participating institutions and provide a general description of support provided to SNFs to implement the program. However, details are lacking on how an external organization, such as a QIO, can recruit SNFs to participate in a QI project involving INTERACT and can facilitate use of the INTERACT tools to decrease hospital readmission rates. We conducted a process evaluation of this pilot QI project with 3 goals: (1) to describe our steps in recruiting SNFs, training SNF staff, and assisting SNF staff with adoption of the INTERACT tools; (2) to determine changes in quality improvement structure and processes, and a specific outcome (30-day all-cause hospital readmission) in the participating SNFs; and (3) to document barriers, successes, and suggested enhancements to our training and technical assistance. Methods Recruitment To identify potential participants, a QIO analyst identified all SNFs within the target community using addresses listed on Medicare’s Nursing Home Compare website9 and calculated 30-day all-cause hospital readmission rates for each facility based on claims data for fee-for-service Medicare beneficiaries for the first 6 months of 2013. Participation criteria included having 40 beds and a 30-day readmission rate of 10% among fee-for-service Medicare beneficiaries during this period. Next, a QIO staff member with an established track record of QI consulting in Connecticut SNFs mailed a recruitment letter to each SNF Administrator or Director of Nursing. This letter was cosigned by the hospital representative who was the designated leader of the community effort to decrease preventable hospital
649
readmissions. The letter emphasized the national focus on reduction of preventable hospital readmissions and Medicare’s strategies of public reporting, financial penalties, and technical assistance from QIOs. Next, 2 other QIO staff members who also had histories of QI consulting to CT SNFs placed follow-up phone calls. These calls were to request a face-to-face meeting with the Administrator, the Director of Nursing, and the Medical Director to discuss the project, answer questions, and obtain leadership commitment to participate. These meetings occurred between July, 1 and November 30, 2013. Quality Improvement Training and Technical Assistance Two QI consultants from the QIO provided individualized training and technical assistance to designated personnel from the participating SNFs from the date of their commitment to the project through June 30, 2014. From July through December, this training focused on use of the INTERACT Hospitalization Tracking Tool. During this phase, the first QI consultant made monthly visits to the SNFs to demonstrate use of the tool, check accuracy of data entry, and review data output. From January through June, the second QI consultant interacted with staff from each SNF to identify opportunities for improvement and to facilitate implementation of the INTERACT tools. All participating SNFs were asked to implement the INTERACT Hospitalization Tracking Tool, the Quality Improvement Tool for Review of Acute Care Transfers, the SituationeBackgroundeAssessmenteRecommendation (SBAR) Communication Form and Progress Note, and the Stop and Watch Tool; additional INTERACT tools were recommended by the QI Consultant based on the specific needs of each institution. Both QI Consultants supplemented their face-to-face visits with phone calls and e-mails, as needed. Data Collection and Analysis During the July through December 2013, training of designated SNF personnel on the INTERACT Hospitalization Tracking tool, the first QI Consultant collected baseline SNF-reported data on resident hospitalizations. From January 1 through June 30 2014, SNF staff collected these data independently and provided them to the QI Consultant on a monthly basis. In December 2013, the second QI Consultant conducted a SNF Needs Assessment that captured administrative information (number of beds, types of units, occupancy rates, for-profit vs nonprofit status, membership in a corporate chain); electronic health record implementation status; tenure of SNF leaders (Administrator, Director of Nursing Services, Medical Director); numbers of physicians, Advanced Practice Registered Nurses; and staffing ratios (minutes per resident day) for registered nurses, certified nursing assistants, and physical therapists; existence of a QI Committee; and use of INTERACT tools. QIO staff recorded quantitative and qualitative data on their interactions with SNF personnel in an electronic database referred to as the Contact Log. They captured dates of interactions, QIO and SNF personnel involved, method of interaction (face-to-face, phone, email, or fax), purpose of interaction (project overview, participation agreement, needs assessment, data collection training, data retrieval, data accuracy assessment, performance data feedback, root cause analysis of resident transfers, staff education, and technical assistance related to INTERACT tool use), duration of contact in minutes, barriers to use of INTERACT tools (competing staff responsibilities, staff availability, staff turnover, clinician resistance to change, inadequate staff communication, technology challenges), use of INTERACT tools prior to the project (Hospitalization Tracking Tool, QI Tool for Review of Acute Care Transfers, Stop and Watch Early Warning Tool, SBAR Communication Form and Progress Note, Facility Capabilities List, Care Paths), and ad hoc notes. In addition to the Contact Log, the QIO team produced internal team meeting minutes that summarized
650
T.P. Meehan Sr. et al. / JAMDA 16 (2015) 648e653
team discussions and problem solving. Finally, the 2 QI Consultants jointly conducted semistructured exit interviews with leaders from each SNF at the conclusion of the project. These interviews elicited SNF leaders’ perspectives on staff use of INTERACT data collection and intervention tools, perceived barriers and benefits to their use, and suggestions for improvement of the QIO training and technical assistance. Both QIO staff members kept independent notes on the interviews; afterwards, they compared notes and reconciled differences in a final written document. Analysis began with calculations of univariate frequencies and simple descriptive statistics for all quantitative data. This included staffing ratios from Nursing Home Compare postings of November 2013, SNF Needs Assessment variables, Contact Log entries, Hospitalization Tracking Tool entries, and quantitative data from the exit interviews with SNF leadership. Changes in QI structure (presence of a QI Committee) and general QI processes (tracking of resident hospitalizations, root cause analysis of resident hospitalizations, and sharing of SNF capabilities with local hospitals) and use of INTERACT tools were assessed by comparing data from the needs assessments and the exit interviews. SNF-specific 30-day all-cause hospital readmission rates were calculated from the Hospitalization Tracking Tool data using the date of admission to the SNF as day zero and using rolling-quarters to compensate for small denominators. Results were graphed in run charts. Statistical significance testing was not done because the numbers were small. Analyses were conducted using Excel, version 2010 (Microsoft, Redmond, WA) and SAS, v 9.2 32-bit (SAS Institute, Inc, Cary, NC). Internal team meeting minutes and SNF leadership exit interview notes were reviewed to validate Contact Log entries and to identify SNF leader suggestions for improvement of QIO-provided training and technical assistance.
Results Recruitment QIO staff mailed recruitment letters to the 15 SNFs within the community that met the selection criteria. Then, QIO personnel made 1e3 phone calls to the Directors of Nursing Services and the Administrators of each SNF to schedule on-site recruitment visits. Leaders at 7 facilities agreed to meet with QIO staff to discuss the project. SNF leaders who were present at the meetings most commonly included the Administrator and the Director of Nursing Services, although others, including the Assistant Director of Nursing Services, the Director of Admissions, the Staff Development Director, and the Infection Preventionist attended at some sites. The Medical Director was not present at any of these initial meetings. Leaders from all 7 SNFs that met with QIO staff agreed to participate in the project. However, both the Administrator and the Director of Nursing Services at 2 SNFs left their positions subsequent to these meetings and these SNFs withdrew from the project. Leaders from the 8 SNFs that did not meet with QIO staff either did not answer the QIO’s initial phone calls or declined to participate, citing lack of time and competing priorities as their reasons. Many of these SNF leaders expressed concern when QIO personnel mentioned that the project would involve use of an electronic hospitalization tracking tool, stating that they were unsure of their staff’s technologic capabilities. In addition, these SNF leaders indicated that they did not see a compelling reason to participate in the project. Table 1 summarizes demographic characteristics and pre-intervention all-cause 30-day readmission rates of the 5 SNFs that participated in the project. Of note, SNF A had an Administrator that had been in place much longer than the other participating facilities; SNF E had much higher staffing ratios; and SNF C had shorter tenures for all of its leadership staff, and
Table 1 Administrative Characteristics of Participating SNFs Characteristic
Beds per unit: Long-term Short-term Dementia care Percent occupancy For-profit Corporate chain Electronic health record Leadership tenure in years: Administrator Director of nursing services Medical director Staffing: Admitting physicians (including medical director) Advanced practice registered nurses Nurse minutes per resident day Certified nursing assistant minutes per resident day Physical therapist minutes per resident day *Pre-intervention 30-day all cause readmission rates (%)
SNF A
B
C
D
E
62 14 95 Yes No Yes
102 30 90 Yes No No
85 30 35 90 Yes Yes No
120 30 95 Yes Yes Yes
53 12 93 Yes No No
30 5 10
5 6 20
