Objective: This study examined the effect of 2 case management delivery models on the organizational outcomes of length of stay (LOS) and payment denials.
Background: Organizational leaders frequently identify case management as a method to manage LOS and demonstrate fiscal responsibility; yet, most case management research is anecdotal, describing specific patient populations or clinical units for outcome measurement.
Methods: Regression analyses and analyses of variance were used to examine the role functions of case managers and the impact on LOS and payment denials. Outcomes were evaluated on medical, surgical, neurology, and cardiology populations across general, intermediate, and intensive levels of care.
Results: Caseload distribution and role definition had a statistically significant impact on LOS and denial management across clincial specialty, units, and levels of care. The role of the case manager predicted 11% of the organizations LOS management.
Conclusion: Case management role definition impacts organizational success measures found in LOS reduction and improved reimbursement. Although only 11% of the LOS was predicted by the role function of care coordinators, altering a single RN role had a statistically significant impact on LOS across clinical specialties and levels of care, demonstrating that indirect care givers have an important impact on organizational success.
In the last decade, significant changes have been made in reimbursement policies on local, state, and national levels. In light of the changes in healthcare reimbursement and increasing costs, there is a need to redefine nursing roles and care delivery systems.1-5 The case management specialty of healthcare was created, in large part, because of the influence of managed care, third-party payers, and reimbursement systems. Care coordination, access to care, delivery of care, resource use, and financial reimbursement are key aspects of the roles fulfilled by case managers to organize care and reduce costs.2,4-6
For healthcare organizations to function successfully, healthcare delivery models must change. The ability to effectively and successfully lead and manage healthcare organizations is hampered by the historic and traditional models of care delivery and organizational management.7-9 There are not enough nurses to fulfill existing roles in the care delivery system, and it is anticipated that the shortage of nursing staff will continue in the coming decades.7,10 The healthcare system is becoming more complicated, creating the imperative for healthcare leaders to identify strategies to address the inefficiencies and uncoordinated healthcare delivery systems through structure and process. A greater understanding of professional roles and role functions, the use of technology to aid communication, and an understanding of financial reimbursement methods will be necessary.
Lack of Evidence
There is a lack of empiric research on the specific structure, role definition, and role functions in case management that leads to organizational success. The burden to validate and verify rationale for using registered nurses (RNs) in job roles not associated with direct patient care has become a paramount responsibility for nursing leaders.7,10 The knowledge, skills, and abilities of RNs are viewed as organizational assets, and organizational leaders who use RNs in case management role functions will need to demonstrate why RN resources should be used for administrative purposes rather than in direct care roles. This study proposes that specific RN case manager role functions and responsibilities can be measured to quantify case manager contributions to overall organizational success.
Models of Case Management
Irrespective of the practice setting or model of case management chosen, central to any case management program are concepts of care coordination; communication and collaboration between healthcare providers, patients, and payers; and attention to the continuum of care for continuity of services provided.4,5 This study was conducted to examine the impact of 2 distinct case management delivery models and the impact on organizational measures of length of stay (LOS) and payment denials. The case manager role functions were completed by RNs in the care coordinator role.
Comparison of Traditional and Full Immersion Delivery Models
The traditional model of case management focuses on discharge planning and utilization reviews for payment authorization by third-party payers. In the traditional case management model, chart review for utilization management functions was expected every 3 days for contracted payers and every 7 days for Medicare and Medicaid. Care coordinator caseloads were between 16 and 28 patients a day.
In addition to the focus on discharge planning and payment authorization, the full immersion model of case management defined additional care coordinator role responsibilities. Documentation expectations changed to include daily chart review; daily electronic documentation in the utilization management and electronic medical record; and daily communication with attending physician staff, consulting physicians, and bedside RN staff for each of the patients in their caseload. Daily interactions between professional staffs were intended to minimize the likelihood of payment denials from untimely communication about treatment plans and interventions to third-party payers. By improving written and verbal communication, the care coordinator could also anticipate potential treatment delays that could be proactively managed. Caseloads for care coordinators were prescribed and identified as 12 to 14 patients on medical and neurology specialty units and 16 to 18 patients on surgical and cardiology units (Figure 1).
|Figure 1. Traditional and full immersion models.|
To address the gap in research on organizationwide impacts of case managers, the overall purpose of the study was the comparison of LOS and payment denial outcomes in the traditional (historical) and full immersion models of case management delivery. The research question for this study was as follows: "what is the relationship between a specific structure, role definition, and role functions in case management delivery that leads to organizational success?" One unique characteristic of this study was the inclusion of cardiac, medical, surgical, and neurology patient populations across general, intermediate, and intensive levels of care.
Human Subjects Protection
Protection of human subjects was undertaken through application and approval of this study by the participating organizations' institutional review board. Data were extracted from the electronic databases and placed into a secure password-protected database. Only the primary investigator and the management engineer completing data extraction had access to the data. Data were organized by unique patient identifiers and patient visits to link information from disparate databases.
This study was completed recognizing that controls over certain variables would not be possible. Existing organizational policies were used for the assignment of patients to a clinical unit during the admission process. Each of the units maintained its clinical specialty, and no changes were made to organizational practices related to the hiring of staff. Physician-admitting practices were not altered.
Patients in this study were adults admitted to and discharged from medical, surgical, neurology, and cardiology units. Data were collected by care coordinators as part of the routine patient care process. Baseline data were collected in the year before the implementation of the full immersion model using the electronic database systems. The utilization management database was used to record utilization management functions, payment denials, and delays in care. Care coordinators also documented assessment and evaluation findings in the electronic medical record database to communicate with healthcare team members.
Causal-comparative designs are used to study relationship between variables to determine differences between groups when the groups are based on preexisting variables or something that cannot be manipulated.10 Causal-comparative designs provide evidence of cause and effect that allows for the study of relationships between independent variables.10,11
In this study, the patients were assigned to different care units based on clinical diagnosis, clinical specialty, and availability of beds based on existing organizational policies. Staff members self-selected the clinical unit they worked on. These practices do not allow for randomization or control. Variables that were not controlled include the assignment of the case manager on a clinical unit, physician practice patterns, and assignment of patients to available beds during the admission process.
Case managers in this study had a bachelor of science in nursing as a requirement to be hired into the care coordinator role. Years of experience ranged from less than 1-year experience as a case manager but no less than 3-year experience as an RN to 7-year experience as a case manager and more than 25 years of nursing experience.
Regression correlation analysis was used to examine the number of utilization reviews conducted by care coordinators for each patient and the percentage of reviews completed to determine if the goal of 100% daily reviews decreased LOS and payment denials. This analysis examined the correlation between LOS and the number of utilization reviews completed and comparisons between the traditional and full immersion delivery models. This represents a testing of the case management delivery models to establish whether the components of the full immersion model would be able to statistically represent differences in LOS across clinical specialties and levels of care.
Analysis of variance (ANOVA) was used to examine payment denials. Denial of payment can be complete or partial. Complete denials for care rendered were based on the payer's determination that acute care was not appropriate or needed. Partial denials were based on payer determinations that treatment delays occurred because of inefficiencies in the organization processes, resulting in specific numbers of days denied payment. If a denial was not identified, the organization was paid in full.
To examine LOS in the traditional model and full immersion model, an ANOVA was conducted. Length of stay was evaluated using baseline data collected in the year before the implementation of the full immersion delivery model on each clinical unit and compared to the LOS post implementation.
In considering whether the differences in LOS could be attributed to specific clinical specialties, a 1-way ANOVA was conducted to examine the LOS and clinical unit specialty. Units included in the study were cardiac intensive care unit, cardiac step down unit, 2 general medical units, medical intensive care unit, neurosurgery progressive care unit, and a general surgical unit.
Results and Findings
Length of Stay
More than 39,000 patients were included in this study (N = 39,109). The average LOS for the 17,669 patients included in the traditional model of case management was 8.69 (SD 10.68) days. The average LOS for the 21,440 patients included in the full immersion model was 7.12 (SD 7.48) days. There were statistically significant differences (P = .00) between groups identified as traditional and full immersion. The LOS was 1.57 days lower in the full immersion sample than in the traditional model sample. This demonstrates a difference in the LOS when case manager role functions, structure, and documentation expectations are defined and measured (Table 1). The variation in LOS was also lower in the full immersion model. The traditional model median LOS was 5.93 days, and the full immersion model median LOS was 5.02 days. This demonstrates less variation in the process to manage LOS.
|Table 1. Length of Stay (LOS) Full Immersion and Traditional Models|
Statistically significant differences were found on all of the clinical units across general, intermediate, and intensive care units (P = .00). This demonstrates that the full immersion model impacts are not isolated to a particular clinical specialty, unit, or level of care (Table 2). This is important because each unit had unique clinical specialties, bedside nurse staffing patterns, years of care coordinator experience, and physician practice patterns. Past research on case management outcomes typically examined a single unit of care or a specific diagnosis requiring care.
|Table 2. Clinical Unit and LOS|
Utilization Reviews Completed
Utilization review documentation was completed in the utilization management database and served as the cornerstone of the differences between the traditional and full immersion models of case management. Documentation in the database served as a proxy that care coordinators reviewed the patient's medical record and had taken steps to communicate findings with other healthcare providers. In addition, as a result of these reviews, steps were taken to facilitate movement of the patient through the hospital system to impact LOS. The utilization review notes are the basis for payer determinations of payment for services rendered. Given an anticipated 5-day LOS, a threshold of 80% of reviews completed was equal to 100%, acknowledging that a utilization review might not be completed on the discharge day (Table 3).
|Table 3. Utilization Management Reviews in Full Immersion and Traditional Models|
In the traditional model, care coordinators were expected to review charts every 3 days for contracted payers and once every 7 days for Medicare and Medicaid patients to arrange discharge plans and authorize payment from third-party payers. In the traditional model, 80% of the utilization reviews were only completed on patients 15% of the time during their hospital stay. Said another way, 85% of the patients did not have a utilization review completed as expected in the care coordinators role description and the organizations contractual requirements.
In the full immersion model, care coordinators were expected to complete daily chart reviews and document in the utilization management database and electronic medical record to expedite care through communication and coordination efforts. Patients cared for on full immersion model units had 80% of the total utilization reviews completed and documented 66% of the time. Patients cared for using the traditional model of case management had 80% of the reviews complete 34% of the time (Table 3).
The components of the full immersion model explain 11% of the LOS value during a hospitalization (R2 = 0.112) (Table 4). The R2 represents the proportion of total variation around the mean, thus the power of the relationship between the components of the model to a straight line (P = .00).
|Table 4. Regression Model|
In examining the dependent variable of LOS and independent variables of year and month of admission and the number of utilization reviews completed per day, there is statistically significance impact of the independent variables on the dependent variable (P = .00) with little collinearity. The variance inflation factor in this regression analysis was 1.038, well below the variance inflation factor value of 10 that demonstrates collinearity.11 This demonstrates that the interaction or overlap between the timing of the admission and the completion of the reviews during each stay had little overlap in influencing LOS (Table 5).The analysis of the pre-post model demonstrated a statistically significant correlation between LOS and the number of reviews completed during the hospital stay (P = .00) (Table 6).
|Table 5. Correlation of Date of Admission and Length of Stay (LOS)|
|Table 6. Correlation Length of Stay (LOS) and Number of Reviews per Day in Stay|
The month and year of the admission were used as a proxy to case manager experience given the role changes for case managers in the full immersion model. Examination of the year and month of the admission did not demonstrate statistical significance between the LOS and the timing of the patients stay (P =.96) (Table 5). This establishes that the experience of the care coordinator completing the review and the date of the admission in relation to the implementation of the full immersion model did not have an impact on LOS.
On the basis of the regression correlation (Table 5), the number of utilization reviews complete was statistically significant to LOS (P = .00). This information coupled with the timing of the admission is important because it illustrates that the completion of the utilization review had an important impact on LOS management but the timing of the admission did not influence LOS.
Using data from the utilization management database, a comparison between payment denials during the traditional and full immersion operations was undertaken. The Healthcare Financial Management Association identifies target denial rate benchmarks of less than 4% of gross revenue for inpatient hospitals.12 In this study, there was a slight increase in denials received during the year that the full immersion model was implemented from 1.08% to 1.28%, well within the benchmark. Only 0.50% of the claims were not paid after appeal. This represents increased revenue capture of $6,553,799.00 and illustrates the effectiveness of the denial management process based on the information in the utilization management database to demonstrate necessity of hospitalization to third-party payers for authorization of payment.
In examination of payment status, a 1-way ANOVA was conducted to determine if the LOS influenced the payment status for services rendered. The LOS did not show statistical significance in whether the claim was paid or denied (P = .302) (Table 7). In addition, the LOS for appealed, denied, or partially denied cases was higher than those paid in full. This demonstrates that LOS was not dependent on the ability of the care coordinator to obtain authorization of payment from third-party payers to continue a hospital stay.
|Table 7. Payment Status (Appealed, Denied, or Partial Denial)|
The first conclusion from this study is that there was a statistically significant reduction in LOS when the caseload staffing patterns and role expectations for the full immersion model were implemented. This outcome validates that caseload staffing patterns and case manager role definition do have an important contribution in organizational LOS management across clinical specialties and levels of care. Patients cared in the full immersion model demonstrated an average 1.57-day reduction in LOS compared with those in the traditional model. Furthermore, the variance in LOS decreased in the full immersion model, suggesting less variation in the process of LOS management. This is an important consideration in outcome measurement and quality improvement principles that validate that reductions in process variation demonstrate greater stability in implemented processes and less likelihood for errors.4,13
A second conclusion from the regression analysis is that role definition and caseload components of the full immersion model explained 11% of the LOS. Although at face value 11% seems to be a low predictive value, in fact, it points to the ability of the full immersion model to statistically represent differences in LOS across clinical specialties and units. Influences such as physician practice patterns, organizational culture, experience of the nursing staff, and payer requirements often become the focus of LOS management.5,8,14 In considering the results of this study, the ability of an organization to impact LOS by changing caseloads and role functions of one group of professionals is significant and, more importantly, manageable when compared with the other areas of focus that typically influence LOS management.
A third conclusion can be made regarding the timing of an admission, care coordinators experience, and time needed to learn a new role before organizational impact would be realized. Organizations often believe that LOS management will improve only slightly at the start of an LOS management program, with improvements being realized over time as the staff becomes more proficient with the change.1,14 This was not demonstrated in this study. The timing of the admission during did not influence the LOS in a statistically significant way. These are important findings for leaders to consider as they plan for the implementation of full immersion delivery model. In part, this finding might be attributed to only changing the care coordinators roles and functions. Physicians and staff nurses were not asked to do anything different in their practice other than talk to the care coordinator daily.
Recommendations for Action
Several areas of organizational operations benefited from this study. The finance department was interested in the findings for 2 reasons. First, the full immersion model was able to demonstrate reduced LOS and increased revenue capture irrespective of the clinical unit for care. Second, although initially the labor costs increased because of reduced caseloads in the full immersion model, the overall reduction in LOS brings balance to the equation of fiscal responsibility.
The quality, compliance, and contracting departments used the findings relative to payment and LOS management to communicate the organization's ability to demonstrate statistically that payment status does not drive discharge or LOS. This is an important information for staff to know because patients and families often believe that they are being discharged sooner because of their insurance plan.
Several recommendations can be made to nursing leaders. Given the nursing shortage, this study provides compelling evidence to closely examine the roles and responsibilities assigned to professional nurses as budgets and staffing plans are created. Organizational leaders are struggling to attract, recruit, and retain professional nurses. Given the shortage of nurses providing care at the bedside, there is often hesitation to take a nurse from a direct care role to complete indirect care functions, and case managers are viewed as indirect care givers in most organizations. This study demonstrates that indirect care givers have an important impact on organizational success highlighted in LOS reduction and reimbursement across clinical specialties and levels of care (Figure 2).
|Figure 2. Outcomes.|
Limitations of the Study
This study was conducted at a large Midwestern hospital that may not reflect experiences of other organizations in the region or country. The accepted expectations for measurable outcomes driven by contractual agreements and local practice norms may have influenced efficiency and effectiveness. In addition, other organizational practices may have influenced LOS, for example, quality and process improvement initiatives. External forces related to reimbursement contracts, changes in patient volumes, and changes in insurance coverage might have influenced LOS and payment denials and were not controlled in this study. Lastly, changes in the mix, education, and years of professional experience of staff were not altered and reflected existing policies in effect during the study period.
Clarity in role function and linkages to organizational outcomes are needed to improve the delivery of healthcare. New paradigms for the delivery of healthcare services will be necessary in organizations, as leaders are confronted with seemingly conflicting pressures surrounding the need to reduce the cost of care while maintaining expected practices for high-quality care with limited human resources. As the country experiences changes in social policy, advances in technology, and persistent demands for high-quality patient care, organizations have the opportunity to radically change care delivery models in a systematic and purposeful way. This study demonstrates that changes in our care delivery models are possible and that increasing communication, collaboration, and coordination of care through the use of RN case managers can have a positive effect on organizational outcomes.
1. Terra S. An evidence-based approach to case management model selection for an acute care facility. Prof Case Manag. 2007;12(3):147-157. [Context Link]
2. Bower K. Patient care management as a global nursing concern. Nurs Adm Q. 2004;28(1):39-43. [Context Link]
3. Clancy T, Delaney C. Complex nursing systems. J Nurs Manag. 2005;13:192-201. [Context Link]
4. Huber D. Leadership and Nursing Care Management. 3rd ed. Philadelphia, PA: W.B. Saunders Company; 2006. [Context Link]
5. Zander K. Nursing case management in the 21st century: intervening where margin meets mission. Nurs Adm Q. 2002;26(5):58-67. [Context Link]
6. Hodgman S. Predictive modeling and outcomes. Prof Case Manag. 2008;13(1):19-23. [Context Link]
7. Buerhaus P. Six-part series on the state of the RN workforce in the United States. Nurs Econ. 2005;23(2):58-60. [Context Link]
8. Storey S, Linden E, Fisher M. Showcasing leadership exemplars to propel professional practice model implementation. J Nurs Adm. 2008;38(3):138-142. [Context Link]
9. Torrens P, Williams S. Managed Care: Restructuring the System. In Introduction to Healthcare in the United States. New York, NY: Delmar Publishing; 2002. [Context Link]
10. Dannenbring G. Descriptive and causal-comparative research designs: Notes. http://www.educ.drake.edu/dannenbring/201and202/ch10_res.html. Accessed September 22, 2008. [Context Link]
11. Aczel A, Sounderpandian J. Complete Business Statistics. 5th ed. Boston, MA: McGraw-Hill; 2002. [Context Link]
12. HFMA self-assessment tool helps to manage denials. December 2, 2004. http://www.hfma.org/publications/know_newsletter/archives/120204.htm. Accessed September 22, 2008. [Context Link]
13. Dedhia N. Six sigma basics. Total Qual Manag. 2005;6(5):567-574. [Context Link]
14. Floyd C. Hospital case management: where are we and how are we doing? Prof Case Manag. 2007;12(3):380-381. [Context Link]