To evaluate the association between patient satisfaction and cost, outcomes, and clinical performance of stroke care.

An ecological study was conducted on all participating hospitals of the Hospital Consumer Assessment of Healthcare Providers and Systems patient survey that reported stroke outcomes.

Patient satisfaction measures were grouped into global, environmental, communication, pain control, staff responsiveness, care transition, and discharge information categories. Linear regression models compared risk-adjusted 30-day mortality, 30-day readmission, inpatient costs, and clinical performances by patient satisfaction.

Global patient satisfaction was negatively associated with risk-adjusted 30-day mortality (beta coefficient [β] = —0.39; standard error [SE], 0.16; 

 = .02) and readmission rates (β = —0.30; SE, 0.11; 

 = .006). Satisfaction with discharge information was positively associated with risk-adjusted 30-day mortality rate (β = 0.70; SE, 0.14; 

 <.001) and negatively associated with readmission rate (β = —0.37; SE, 0.09; 

 <.001). Satisfaction with discharge information were positively associated with inpatient management (β = 1.67; SE, 0.43; 

 <.001) and secondary care performance (β = 1.82; SE, 0.47; 

 <.001). The average cost among most satisfied hospitals was $6785, 7.3% higher than that among least satisfied hospitals ($6324). Hospitals with the highest environment satisfaction rating had 7% higher costs compared with the least satisfied hospitals.

Global patient satisfaction was positively associated with the quality of stroke care; however, improvements in patient satisfaction were linked to higher stroke care costs. In addition, patient satisfaction with discharge information was linked to worse outcomes. As a result, patient satisfaction should be used with caution as a quality indicator for stroke care. 

Although previous studies showed that higher patient satisfaction was associated with reduced readmission and mortality rates, it is still not clear whether patient satisfaction can be a quality indicator of stroke care. 

Higher overall patient satisfaction with stroke care was associated with lower readmission and mortality rates. 

Higher patient satisfaction with stroke care discharge information was associated with lower readmission rate and better stroke care performance but higher mortality rates. 

Higher patient satisfaction with stroke care was accompanied by higher costs.

With an increasing emphasis on curbing costs while improving healthcare, patient satisfaction has become an essential quality indicator. Patient satisfaction measures are an important component of the Affordable Care Act’s financial incentives for hospitals to provide better quality of care.

 In the Donabedian framework for examining health service and quality, healthcare quality comprises the dimensions of structure, process, and outcomes.

 Fortunately, the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey provides an opportunity to capture the patient experience with a variety of quality categories, including environment, communication, pain control, responsiveness, and care continuity in hospitals, allowing us to further explore the role of patient satisfaction in distinct quality aspects.

The HCAHPS measures were built into financial incentives. In 2012, CMS began withholding 1% of reimbursements from hospitals as part of its value-based purchasing (VBP) program. Hospitals are able to earn the payment back by realizing high patient satisfaction scores and care quality standards. Tying patient satisfaction to hospital reimbursements remains controversial, as more satisfied patients do not necessarily receive better medical care or experience superior health outcomes. Moreover, there is a controversy over the efficacy and reliability of patient satisfaction metrics. There is a concern that unnecessary costs may increase in the pursuit of higher scores on these metrics, because doctors may turn to discretional services when their payments are associated with patient satisfaction.

The controversies urged researchers to examine the association between patient satisfaction and healthcare quality in distinct classes of diseases. Stroke is a class of life-threatening cerebrovascular disease that caused 129,000 deaths and cost $20.2 billion in 2010 in the United States.

 If boosting patient satisfaction can improve stroke care quality without significantly increasing the cost, patients and hospitals will benefit.

So far, empirical studies have studied the association between patient satisfaction and healthcare cost and quality in various diseases. Evidence suggests that hospitals with higher patient satisfaction scores have lower readmission rates for acute myocardial infarction, heart failure, and pneumonia,

 and patient satisfaction has also been shown to improve guideline adherence and mortality in acute myocardial infarction.

 For surgical care, higher patient satisfaction saw shorter inpatient stays and lower readmission and mortality rates.

 However, higher patient satisfaction was also associated with higher inpatient costs and postdischarge mortality rates on the individual patient level.

 Moreover, most existing studies have assumed all satisfaction categories to have a similar effect as global satisfaction on care quality. It is well documented that higher global patient satisfaction was associated with better care qualities, yet how the specific aspects of patient satisfaction relate to other structure and process quality measures is still unknown.

Therefore, we explored the relationship between patient satisfaction and cost, clinical performance, and health outcomes in stroke care at the hospital level. We examined the following research questions: first, whether global satisfaction was associated with better stroke care quality; second, whether the components of satisfaction were independently associated with stroke care performance and outcomes; and finally, we assessed whether improving patient satisfaction resulted in higher stroke care cost and what satisfaction categories accounted for the differences in costs.

We obtained hospital-level patient satisfaction measures from the 2015 HCAHPS patient survey database.

 The average inpatient charges for hospitalized patients who had a stroke came from the 2013 CMS Medicare Provider Cost and Payment Data inpatient file. Outcome measures were selected from the CMS Hospital Compare database, including risk-adjusted 30-day mortality rates and risk-adjusted 30-day unplanned readmission rates. This database was updated between 2014 and 2015 at the time of the study. The clinical performance measures for timely and effective stroke care also came from Hospital Compare database (

 We further augmented our data with hospital characteristics from the fiscal year 2016 Inpatient Prospective Payment System final rule file.

 We obtained the US Census geographic region from the 2014 Area Health Resource File.

Hospital-level mortality and unplanned readmission rates were already adjusted by patient demographic characteristics and CMS condition categories (eAppendix Table 1). Thus, these risk-standardized measures allowed us to compare hospitals with different patient demographics and health status distribution.

Cost was defined as the cost of care for patients who have had a stroke without major comorbidity. Specifically, we estimated cost by multiplying the charges for Diagnosis-Related Group (DRG) 066 by hospital-specific Cost-to-Charge Ratios (CCRs). We used the CCRs from the fiscal year 2016 file for cost in 2013 because of the 3-year lag period in cost data collection.

According to a previous classification of stroke care performance measure, we constructed 3 performance domains, including acute treatment, inpatient management performance, and secondary prevention performance.

 We generated the domain performance scores (range = 0-100) by averaging the scores of measures that belonged to the same domain (eAppendix Table 1). We used 6 items from a CMS measurement of timely and effective stroke care to indicate clinical performance of stroke care (eAppendix Table 1).

 “Intravenous thrombolytic therapy” and “patients with atrial fibrillation receiving anticoagulation therapy” were dropped from analysis because of an excessive missing rate (>50%). 

The data from the HCAHPS survey provided us with standardized measures of patient satisfaction with hospital care for all patients at the hospital level and were adjusted for patient-mix and survey mode.

 CMS classified the HCAHPS survey questions into 2 global items, 2 individual items, and 7 composite measures of patient satisfaction.

 Patient satisfaction categories in our analyses were selected from this preexisting classification.

The global items were an overall rating of the hospital and the patient’s likelihood of recommending the hospital to friends and family. Because these 2 global measures were highly correlated, we only used the recommendation measure for analyses. The patient satisfaction score is the percentage of patients with the highest possible rating on the metric. We averaged the scores in the same category, and then divided these scores into quartiles (eAppendix Table 1).

Multiple hospital characteristics were included as covariates in the analyses. Hospitals’ geographic characteristics included whether they were located in an urban or rural setting and their census region locations (northeast, midwest, west, or south). Teaching status was determined by whether a hospital received a teaching hospital payment adjustment from CMS. Hospital ownership was classified as public, for-profit private, nonprofit private, or other. We classified hospitals as small, medium, or large according to the number of beds, teaching status, and region, following the definition used by the Healthcare Cost and Utilization Project.

 Disproportionate share hospitals (DSH) were defined by CMS as the sum of the percentage of Medicare and Medicaid days that were attributed to low-income patients who were Supplementary Security Income eligible. We also determined from these data whether or not a hospital participated in a stroke care registry.

We first described the distribution of hospital characteristics in those that reported stroke outcome measures or cost information. Categorical data were presented as counts with percentages and continuous variables were presented as medians with interquartile ranges.

We estimated coefficients (β) and 95% CIs for the association between patient satisfaction and cost, 30-day mortality rate, 30-day readmission rate, and stroke care performance for all hospitals reporting stroke outcome data, using multiple linear regression models. The regression model for cost was performed on hospitals that reported cost information for DRG 066. We applied a log transformation to the cost data. Thus, the β coefficients yielded a percentage difference in cost compared with the reference group. The percentage change of the cost was calculated as (

Regression models were adjusted for hospital size, ownership, teaching status, region, rural location, DSH percent, and stroke registry status. For models regarding stroke outcomes, we added stroke performance measures to the adjustment set to examine whether patient satisfaction measures were independently associated with stroke outcomes. To make the results more generalizable to all strokes, even those accompanied by complications and comorbidities, we did a sensitivity analysis combining all stroke DRGs (65 and 66). All statistical analyses were performed using SAS version 9.4 (SAS Institute; Cary, North Carolina).

We identified 2530 hospitals that reported 30-day mortality rates and 30-day readmission rates for stroke and 1422 hospitals that reported an average cost for stroke without major comorbidities and conditions in fiscal year 2013.

 describes the characteristics of selected hospitals. In hospitals that reported stroke outcomes, 15.1% were publicly owned, 17.1% were located in the northwestern region, 22.2% were small, and 22.2% were located in a rural area. The median 30-day mortality and readmission rate was 14.7 and 12.7 per 100 patients, respectively (interquartile ranges, 2.1 and 1.4, respectively). The median scores for acute treatment, inpatient management, and secondary prevention were 99.0, 98.5, and 97.6 (interquartile ranges, 2.0, 3.5, and 5.1).

We found that the global satisfaction measure (“recommendation of hospital”) was related to 30-day outcomes. The 30-day mortality rate comparing the highest quartile to the lowest quartile of global satisfaction was significantly lower (β = 0.39 per 100 patients) (

). Similarly, the 30-day readmission rate was also lower in the highest quartile of global satisfaction compared with the lowest quartile (β = 0.34 per 100 patients) (Table 2). The 

(a and b) demonstrated a dose-dependent effect for 30-day readmission across quartiles of global satisfaction, with a negative association between satisfaction and 30-day mortality in the upper 3 quartiles.

Hospitals that ranked the highest in satisfaction with hospital environment had a higher 30-day readmission rate by 0.27 per 100 patients compared with the least satisfied quartile. However, 30-day mortality was not associated with satisfaction with hospital environment. The highest quartile of satisfaction with pain control was associated with lower 30-day mortality rate (β = —0.57 per 100 patients) (Table 2). In contrast, we observed an increase in 30-day mortality rate of 0.70 per 100 patients in the highest quartile of satisfaction with discharge information relative to the lowest quartile. However, readmission rates were lower in hospitals with the highest, versus the lowest, satisfaction with discharge information (β = –0.37 per 100 patients) (Table 2). The full regression results can be found in the 

The average cost in hospitals with the highest global satisfaction was $6785, 7.3% higher than the least satisfied hospitals ($6324) (

[c]). There is a decreasing trend of cost within the first 3 quartiles of satisfaction, but with no statistical or clinical significance. After adjusting for hospital-level covariates, the cost in hospitals with highest global satisfaction was still 8% higher than the least satisfied hospitals. In addition, compared with hospitals in the least satisfied quartile of environment, the highest quartile had 7% higher costs. Meanwhile, satisfaction with communication, pain control, care transition, and discharge information was not associated with cost (Table 2).

), the hospitals with the highest global satisfaction cost 8% higher than the least satisfied hospitals, even after combining the cost of all patients who had a stroke (with and without major complications).

Patient Satisfaction and Stroke Care Performance

No association was found between global satisfaction and acute treatment, inpatient management, or secondary prevention performance. Hospitals in the highest quartile of discharge information satisfaction had 1.62 (

 <.001) higher scores, on average, with regard to inpatient management and secondary prevention than hospitals in the lowest quartile (

). Full regression results can be found in 

Understanding whether patient satisfaction with hospital care is consistent with healthcare quality is critical to current CMS reimbursement policies based on a VBP system.

 Our findings support the hypothesis that global patient satisfaction can represent the quality of care that hospitals provide. Specifically, we found that the risk-adjusted 30-day mortality rate and readmission rate among patients who have had a stroke decreased, with an increase in global patient satisfaction from the lowest to the highest quartile. Linking patient satisfaction to healthcare quality, our study results suggest that the global patient satisfaction measure is useful for assessing value-based care in the context of stroke.

Our investigation into each satisfaction dimension also challenged the conventional assumptions in existing studies that all satisfaction dimensions have similar effects on care quality. Previous studies found that improving the quality of care continuity, such as care transition, patient education, and discharge planning, was effective in containing unplanned readmission rates.

 We observed that patient satisfaction with discharge information was associated with a reduced readmission rate. Surprisingly, we also observed higher satisfaction with discharge information associated with higher 30-day mortality rates. A previous systematic review of various diseases showed that discharge planning did not help reduce the mortality rate,

 leading us to reconsider the role of discharge information in quality improvement.

The results of a recent study show that the hospital-level mortality rates were not associated with the readmission rates, proposing the hypothesis that the 30-day readmission rate and mortality rate are conveying different information about healthcare quality.

 It should be noted that the 30-day mortality rate was calculated from the moment of inpatient admission and was largely affected by in-hospital care, whereas 30-day readmission rates were calculated only after discharge. There was even an article that casted doubt on the validity of readmission rates, suggesting that it is more an accountability measure than a quality measure.

 Thus, patient experience measures that are found to be associated with readmission reduction do not necessarily also help to reduce mortality rate.

It is possible that resource abundant hospitals have good quality and accountability and generally have better patient experience. However, hospitals that have to spend substantial efforts on improving patient education, discharge planning, and outpatient support may have higher patient satisfaction and lower readmission rates; however, such efforts could drive stroke mortality in the opposite direction. In our analysis, patient satisfaction with discharge information was linked to better inpatient management and secondary prophylaxis performance, while the increased performance was not associated with better health outcomes. Thus, the incentive to achieve premium patient experience may be unfair to hospitals that cannot spare enough resources to support all patient satisfaction—promoting activities.

VBP programs are primarily aimed at improving healthcare quality while reducing cost, yet we found that hospitals in the highest quartile of satisfaction provided care that costs 6.9% more than hospitals in the lowest quartile, suggesting that improving patient satisfaction does not help to mitigate inpatient costs. Hospitals that take a large number of patients who have suffered a stroke will spend substantially more money to achieve higher satisfaction. This is an interesting finding, because we observed that the cost rose faster than any improvements in outcomes. The observed improvements in mortality and readmission rates were 3.1% and 2.7%, respectively, which were similar in size to a previous study of patients with acute myocardial infarction, heart failure, and pneumonia.

 The results of a previous study showed that higher patient satisfaction was linked to higher intensity of inpatient care and drug prescriptions.

 Yet, these procedures and prescriptions were not necessarily beneficial to patients. This finding suggests that hospitals that invested more money to improve patient satisfaction generated a minimum or even a negative return in stroke care outcome.

There are several limitations to our study. First, as it is based on CMS data, our results may not be appropriate to generalize to all hospitals because the participating hospitals may be systematically different, in terms of hospital characteristics, compared with non-participating hospitals. Another limitation is the ecological design of our study. Our results can only explain the association at the hospital level, not at the patient level. Although we have adjusted for hospital size, teaching status, location, region, and disproportionate share, unobserved factors may still confound our results. For example, hospitals that discharge more patients to their home, rather than a skilled nursing facility, may have higher patient satisfaction and lower readmission rates. Nevertheless, lacking stroke severity information may not be a major confounder because the health outcome measures were risk adjusted in terms of case mix, hospital clustering effect, and comorbidities of patients.

 The risk adjustment may also explain the lack of variability of mortality and readmission rate among hospitals in our sample. Furthermore, our study cannot determine the direction of causality. HCAHPS surveys were conducted from 48 hours to 6 weeks after discharge. It could be possible that some patients reported low satisfaction because they experienced unfavorable health outcomes during the time after discharge. Finally, without an indicator of stroke severity, we can only provide the evidence of association, instead of causality, between patient satisfaction and stroke care cost.

Despite the limitations, our study also has some strengths. To extend our ability to extrapolate our conclusions to a broader definition of stroke cases, we conducted a sensitivity analysis. In those analyses, global patient satisfaction remained associated with an increased cost of care. In addition, we used both CCR and hospital charge to estimate cost. Specifically, we estimated cost by multiplying the charges for DRG 066 by hospital-specific cost-to-charge ratios. Although we do not have information of the real costs that were incurred, previous researchers have verified that using the CCR produces estimates closest to providers’ actual costs.

 There are many other ways to estimate cost, such as using the CMS payments or hospital charges; however, these methods are also subject to limitations. The CMS payments were the reimbursements to the hospitals after adjusting the costs by regional, hospital, and individual characteristics according to published policies.

 With hospital-level aggregated data, we were not able to disentangle policy adjustment and the real cost from the payments. The charges were the list prices marked by the hospitals, which were usually much higher than the real cost.

 Thus, standardizing charges by CCR was the best approximation of cost in this context.

Articles

WEB EXCLUSIVE

Global patient satisfaction was positively associated with quality of stroke care and higher discharge information satisfaction may be linked to worse outcomes. Additionally, improvements in satisfactions were linked to higher costs.

Is Higher Patient Satisfaction Associated With Better Stroke Outcomes?

has been adopted by many health systems to improve care and decrease costs through coordination of care across levels. At our academic medical center, several care management programs were developed under separate management units, including an inpatient-based program for all patients and an outpatient-based program for complex, high-utilizing patients. To bridge administrative silos between programs, we examined longitudinal care experiences of hospitalized complex patients to identify process and communication gaps, drive organizational change, and improve care. 

 This descriptive study analyzed the care experiences of 17 high-utilizing patients within the authors’ health system.

 Chart audits were conducted for 17 high-utilizing patients with 30-day hospital readmissions during 2013. Clinical and social characteristics were reviewed for patterns of care potentially driving readmissions. 

 Patients had heterogeneous social factors and medical, psychological, and cognitive conditions. Care management interventions apparently associated with improvements in health and reductions in hospitalization utilization

included movement to supervised living, depression treatment, and achievement of sobriety. Monthly case management meetings were restructured to include inpatient, outpatient, ambulatory care, and emergency department care managers to improve communication and process. During 2014 and 2015, hospital readmission rates were overall unchanged compared with base year 2013 among a comparable cohort of high-utilizing patients. 

Joint review of clinical characteristics and longitudinal care experiences of high-utilizing, complex patients facilitated movement of historically siloed care management programs from their focus along administrative lines to a longitudinal, patient-centered focus. Decreasing readmission rates among complex patients may require direct linkages with social, mental health, and substance use services outside the healthcare system and improved discharge planning.

To bridge administrative silos between care management programs, we examined longitudinal care experiences of hospitalized complex patients to identify process and communication gaps, drive organizational change, and improve care. 

Charts of 17 high-utilizing patients with heterogeneous medical and psychological conditions were reviewed for clinical and social characteristics and patterns of care potentially driving readmissions. 

Care management interventions apparently associated with improvements in health and reductions in hospitalization utilization included movement to supervised living, depression treatment, and achievement of sobriety. 

We are restructuring our monthly case management meetings to ensure simultaneous participation of case managers from inpatient, outpatient, ambulatory care, and emergency department service lines to eliminate communication and process gaps and better serve the wide range of clinical and social service needs of complex high-utilizing patients.

Thirty-day hospital readmission rates are high among individuals with complex medical and psychosocial needs.

 Although findings of randomized controlled trials (RCTs) have shown care transition interventions can reduce hospital readmissions in older adults

 populations with multiple comorbidities and mental illness or substance abuse present special challenges. Some programs report success in improving services for this population,

 but we identified no RCTs showing reductions in readmission. The purpose of this paper is to describe the clinical characteristics and health service needs of readmitted patients with complex medical and psychosocial needs, with implications for linkages between inpatient and outpatient care management programs.

At our academic medical center, a number of care management programs focused on diseases or settings were developed to improve care transitions and reduce readmissions. Most evolved through separate administrative units without establishing communication or collaboration processes across units or levels of care. As these programs grew, one vexing gap in coordination occurred between inpatient and outpatient programs. This was highlighted in 2010 when changes to inpatient discharge planning were undertaken and renamed care management without coordinating with the outpatient care managers caring for many of the most complex patients. Therefore, we wished to build coordinated care management protocols between inpatient care management nurses and social workers and the outpatient care management social workers working with complex patients. This program, our Complex Care Management Program (CCMP), is primarily staffed by social workers due to patients’ behavioral and socioeconomic needs. To facilitate this process, we applied the Lean process from Toyota,

 which asserts that problems are best analyzed and addressed by the individuals closest to the data from the point of delivery. Specifically, we undertook a patient-level analysis of a cohort of complex patients with 30-day readmissions.

The CCMP provides care management to individuals with significant challenges in several of 5 domains based on the 

Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition

 classification system: major psychiatric conditions, behavioral and substance use disorders, active medical conditions, lack of social support, and inadequate physical resources (eg, precarious housing or food, limited transportation, or multiple socioeconomic barriers). At the time of this study in 2013, patients were enrolled in CCMP primarily through record review of Medicaid- and Medicare-enrolled and uninsured patients discharged from inpatient and emergency department (ED) services. A complex care manager conducted a phone assessment of each candidate patient, using questions covering the 5 domains with additional tailored follow-up questions, such as their understanding of and ability to obtain medications and keep appointments.

 Most received only this phone call, but during the study period, approximately 10% to 15% of patients assessed were identified as complex based on care managers’ qualitative judgment—usually indicating challenges in at least 3 of the 5 domains&mdash;and, if the patient was not known to CCMP care managers, were enrolled for ongoing care management. Incorporating key elements of successful care management programs for complex patients, CCMP care managers coordinate services for patients, accompany them to primary care and specialty appointments,

 and help to identify and reduce socioeconomic barriers to treatment.

 The usual caseload of complex patients per care manager is 25 to 35 patients.

 The CCMP serves approximately 450 patients annually.

We conducted a descriptive longitudinal analysis of patient experiences of a consecutive cohort of complex patients across levels of care at our institution. Seventeen CCMP patients with 1 or more 30-day readmissions during the first 3 quarters of 2013 were audited, including all with 30-day readmissions in 2 (n = 7) or 3 (n = 1) of the first 3 quarters. In addition, 7 patients with at least 1 readmission from the third quarter (the most recent data available) and 1 patient from each of the previous 2 quarters were audited (n = 9). Most were readmitted multiple times during the study period.

A nurse practitioner experienced in working with similar populations but unassociated with the medical center analyzed the data, identifying patients’ clinical conditions, presenting complaints at each hospitalization, types of healthcare professionals caring for the patient, and services provided during and between hospitalizations. Clinical judgment was used to identify factors potentially influencing readmission. Factors were qualitatively organized in categories mirroring the 5 domains used as enrollment criteria for CCMP. 

Patients ranged in age from 18 to 61 years; 7 were women and 10 were men. All but 1 had income below the federal poverty line, and most had Medicaid or Medicare or were uninsured. All patients had 2 or more chronic medical problems, multiple socioeconomic barriers, and/or were nonadherent to treatment. Many used food assistance programs and had limited social support, strained family relationships, and transportation problems (sometimes complicated by assistive devices, such as walkers). Several patients lived in group homes; 8 had precarious housing situations (

). Frequent housing changes complicated patient management and follow-up for medical and psychological conditions.

The most common chronic medical diagnoses were diabetes, hypertension, angina, chronic obstructive pulmonary disease (COPD), end-stage renal disease, and congestive heart failure. Two patients died during the 6-month audit period as a result of medical complications. Inpatient admissions over the study period ranged from 2 to 5 per patient (mean = 3.3). Common admission diagnoses were alcohol or substance toxicity/withdrawal, respiratory disease (asthma/COPD), chest pain, and psychological conditions (depression, suicidality, and psychosis). With the exception of alcohol/substance use—related admissions, readmissions were often for a different diagnosis than the previous admission, and no simple pattern emerged. 

Sixteen patients had psychiatric or psychological conditions; 1 had cognitive impairment. The most common psychological condition was depression, followed by anxiety, schizophrenia, and bipolar disorder. These conditions sometimes complicated efforts to manage medical conditions. In 2 patients, hallucinations and psychiatric medication adjustments caused behavioral changes that complicated self-management and access to healthcare. In others, depression or anxiety interfered with the ability to manage their conditions but was not considered severe enough to justify intervention by county Community Mental Health psychiatric providers, resulting in treatment delay. In another case, a patient was eligible for a program in which his psychiatric medication was monitored; however, no medication monitoring was available for his medical condition and he was re-hospitalized for that condition. 

Eight patients had substance use disorders, with alcohol dependence being the most common. Most readmissions in this group were related to substance use. Two were lost to follow-up during the study period.

Care Coordination Gaps and Institutional Response

Patients accessed health, behavioral, and social services at the community level, as well as multiple levels within the health system, and communication gaps between these services further challenged their care. In some cases, these challenges appeared to be factors in readmission, either through missed communication or the inability to effectively treat the patient’s medical or psychological conditions. In addition to primary care, emergency, and inpatient services, most patients saw 1 or more medical specialists and used community services related to housing, transportation, food, medications, health insurance, or mental health. 

One important challenge was that the electronic health records (EHRs) used in the community, outpatient, and inpatient services were not accessible by other agencies or different levels in the health system. Although CCMP personnel had access to health system records, health system personnel did not have access to CCMP records and neither group had access to community service records, possibly contributing to missed communication between levels of care. In 2 patients transferred to skilled nursing facilities, for example, medications for chronic medical/psychiatric conditions were not given, leading to 1 readmission and 1 loss to follow-up. In others, late or incomplete communication about laboratory results, medication dosages, or specialty follow-up preceded readmission. In addition, no system existed to communicate patient hospitalizations to CCMP staff and inpatient care managers were not always aware of CCMP resources. 

Identifying and acting on these communication challenges offered an early way for the health system to respond and potentially reduce readmissions. Administrators and care managers from the inpatient and CCMP programs reviewed the care experiences and communication gaps described in the audit at meetings throughout 2014. Timely communication among the patient, CCMP, and healthcare personnel was identified as a critical need related to readmission and coordination of posthospital services and referrals. As a result, program reform shifted toward improving communication, and patient-specific care planning was initiated between the inpatient care managers and the CCMP case managers who coordinated posthospital services. 

In early 2015, monthly meetings between inpatient care managers and CCMP care managers grew to include primary care—based care managers. Through most of 2015, meetings focused on communication and documentation protocols to enhance the visibility of care planning notes among inpatient, CCMP, and outpatient care managers, for example, by creating a new, more recognizable label for EHR encounters (eg, “complex care management” rather than the more generic “telephone note”). Early recognition of CCMP patients by inpatient care managers was improved through use of a flagging system in the EHR that indicated the patient’s enrollment in the CCMP and contact information for the CCMP care manager, allowing them to be invited to inpatient care planning meetings. In 2016, outpatient social workers began to participate in these care planning meetings. Bolstered by relationships established at these meetings, joint assessment and care planning are now routine and referrals from inpatient care managers to those from the CCMP have increased. The value of the CCMP in behavioral management planning and bridging between inpatient settings and primary care also motivated increased communication between inpatient and ED care managers and CCMP; these groups now meet regularly for patient care planning and process improvement. Use of a consistent EHR across all levels of care in the health system has also improved communication. 

In addition to communication within the health system, communication is critical between the health system care management programs and community services providing patients with ongoing treatment or safer hospital transitions. Although incompatible EHRs remain, it was clear that case managers in both the health system and community worked to identify services relevant to patients’ physical and psychological conditions. In some cases, medications or treatments were obtained after substantial communication with insurance companies, specialty care providers, or pharmacies. In others, identifying relevant nonmedical community services improved patients’ self-management of their chronic conditions. For example, 2 patients improved after low-cost counseling was identified and provided for depression and anxiety, 2 after movement to group homes, and 1 after becoming sober. Regular meetings are also held among CCMP personnel and community agencies.

Improving care planning around behavioral conditions is recognized as a priority but has been more difficult to address. Barriers include the historical lack of involvement by Community Mental Health providers when patients undergo a medical admission, even when care of the medical condition is complicated by the patient’s psychological condition. A work group including University of Michigan Psychiatry, the CCMP, inpatient care management, and Community Mental Health is being convened to facilitate communication among these settings while patients are in the ED or inpatient services. 

Despite improvements in early recognition, patient-specific root-cause analysis, and coordinated care planning, 30-day hospital readmission rates for all hospitalizations among CCMP clients have been relatively stable over time (

). Patient-specific readmission rates showed similar results (not shown). 

We undertook a patient-centered approach to begin to develop process solutions to reduce readmissions and support patients as they transition from inpatient to community settings. We used a readily available resource, longitudinal patient records, to examine patient experiences across levels of care and community services. Coordination gaps identified were used to develop improved communication processes among the multiple settings providing services to this population. Other factors contributing to readmissions were varied and included the complexity and severity of medical conditions, psychological and socioeconomic conditions that complicated disease management and care access, and substance use. This is consistent with prior work finding that lack of social support or resources, a history of substance abuse and/or mental illness, and difficulty obtaining medication or transportation impacted posthospital transitions of adults,

Several factors may explain the lack of effect on hospital readmission rates among CCMP patients. Overall, the portion of preventable hospital costs among high-risk populations has been estimated at just 6%.

 In addition, studies demonstrating a relationship between effective discharge planning and decreased readmission rates have largely not been conducted in this population.

 Decreasing avoidable hospitalizations in this population will likely require larger system reforms that address psychosocial determinants of health, the supply of behavioral health providers, and improved coordination of social service organizations with primary care. 

Although the system modifications we have begun have not resulted in measurable reductions in readmission rates, important changes have occurred. By using a review of patient experiences as a starting point rather than administrative protocols, new and productive conversations across units occurred. Previously siloed care managers in inpatient and ED settings now meet regularly with outpatient care managers to coordinate care for individual patients. Care management documentation protocols have been shared and made more visible in the EHR across units. Perhaps most importantly, working relationships among staff and managers from different administrative units have developed, fostering interest in continuing to find new ways to promote patient-centered, rather than unit-centered, care management at all levels, including care managers, program management, and senior leadership.

Only qualitative analysis was conducted; future research should include quantitative analysis of other factors such as length of stay and cost. In addition, the sample size was small, limiting generalizability. Finally, only 3 quarters were examined. However, consistent themes emerged across patients, suggesting that relevant factors had been identified for this population.

Effective care management programs for populations with medical, psychological, and social issues require collaboration across levels and disciplines, particularly with mental health and substance use providers,

 identification of patients at high risk of readmission,

 involvement of interdisciplinary teams in ongoing patient care,

 and building of trusting relationships between patients and care providers.

 However, development of horizontal organizational reforms that require collaboration across administrative units can be challenging in organizations where these units have historically operated independently. Using actual patients’ longitudinal experiences across inpatient and outpatient settings can help break down barriers to collaboration and joint planning, and foster trust and cultural transformation to promote further reform.

 University of Michigan (NG, CD, BCW), Ann Arbor, MI; Complex Care Management Program, University of Michigan (DF, BCW), Ann Arbor, MI.

 The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article. 

The authors audited a series of complex patients’ records longitudinally across their institution’s existing care management programs to improve the coordinated functioning of these programs.

Improving Care Transitions: Complex High-Utilizing Patient Experiences Guide Reform

To develop a model of the option value a therapy provides by enabling patients to live to see subsequent innovations and to apply the model to the case of nivolumab in renal cell carcinoma (RCC) and non—small cell lung cancer (NSCLC).

A model of the option value of nivolumab in RCC and NSCLC was developed and estimated.

Data from the Surveillance, Epidemiology, and End Results (SEER) cancer registry and published clinical trial results were used to estimate survival curves for metastatic cancer patients with RCC, squamous NSCLC, or nonsquamous NSCLC. To estimate the conventional value of nivolumab, survival with the pre-nivolumab standard of care was compared with survival with nivolumab assuming no future innovation. To estimate the option value of nivolumab, long-term survival trends in RCC and squamous and nonsquamous NSCLC were measured in SEER to forecast mortality improvements that nivolumab patients may live to see.

Compared with the previous standard of care, nivolumab extended life expectancy by 6.3 months in RCC, 7.5 months in squamous NSCLC, and 4.5 months in nonsquamous NSCLC, according to conventional methods. Accounting for expected future mortality trends, nivolumab patients are likely to gain an additional 1.2 months in RCC, 0.4 months in squamous NSCLC, and 0.5 months in nonsquamous NSCLC. These option values correspond to 18%, 5%, and 10% of the conventional value of nivolumab, respectively.

Option value is important when valuing therapies like nivolumab that extend life in a rapidly evolving area of care. 

Therapies that enable patients to live to see further innovations in care have option value. 

Option value raises the conventionally estimated value of nivolumab to patients with metastatic cancer by 18% in renal cell carcinoma, 5% in squamous non—small cell lung cancer (NSCLC), and 10% in nonsquamous NSCLC. 

Option value is particularly important in disease areas where there is rapid innovation. In such areas, payers and providers should consider option value when gauging the value of new therapies.

Cancer is a leading cause of death in the United States,

 with rising prevalence. In 1995, there were 7.1 million individuals in the US who had ever been diagnosed with cancer (2.7% of the population)

; by 2012, there were 13.4 million such individuals (4.3%).

 Although cancer mortality rates are declining,

 prevalence is expected to increase because of population aging,

 advances in cardiovascular care, earlier cancer detection, and life-extending treatment.

Increasing cancer prevalence translates to greater demand for oncology care, with implications for payers. Given growing pressures on payers’ budgets and debate over the cost of cancer treatment,

 value is an increasingly important consideration in coverage and access decisions. Moreover, the pressures on payers will increase as cancer prevalence rises, further reinforcing the importance of value.

As value has grown in importance in assessing cancer care, organizations have offered recommendations on measuring value. For example, the National Comprehensive Cancer Network and the American Society of Clinical Oncology recently published frameworks for considering a therapy’s efficacy and cost to determine its value.

 The Institute of Clinical and Economic Review, an independent health technology assessment (HTA) organization, uses cost-effectiveness analysis (CEA) and budget impact analysis to measure therapies’ value,

Traditional methods of CEA measure the value of an innovation by comparing benefits such as survival gains and improved quality of life with costs, assuming no other health technology improvements besides the one in question.

 However, such methods ignore aspects of therapies that matter to patients, like the value of hope, insurance value to healthy individuals, and the increased value people place on life near its end.

Such components of value can be difficult to quantify in HTA because they are subjective valuations based on unobserved patient preferences. However, 1 relevant component of value is rooted in real-world survival data, the “option value” of therapy. Option value is the benefit a therapy provides patients by enabling them to survive to the next innovation. For example, in cancer care, any survival gains today hold additional value because they increase the likelihood that a patient may live to access even more effective treatment in the future. This concept is particularly important in disease areas with traditionally poor outcomes and rapid innovation. 

Although metastatic non—small cell lung cancer (NSCLC) and metastatic renal cell carcinoma (RCC) have had low survival rates,

 innovation to treat both cancers is currently rapid. A number of breakthrough targeted therapies and immuno-oncology (IO) therapies, which harness the immune system to fight cancer, are available, such as nivolumab, pembrolizumab, osimertinib, and ceritinib for NSCLC and nivolumab and lenvatinib for RCC.

 These therapies are improving 1 or more clinical endpoints, such as survival rates for cancers with traditionally poor survival odds.

 Furthermore, the pipelines for both tumors are active.

Given the proliferation of treatment options and the prospect of future innovation, payers, policymakers, clinicians, and patients require multiple criteria to make care-related decisions.

 A therapy’s option value accounts for the future benefits of a life-extending therapy in addition to its immediate survival effects. With that in mind, we looked at the option value of nivolumab, the first IO therapy given breakthrough status and approved by the FDA for NSCLC and RCC.

 NSCLC is histologically divided into squamous and nonsquamous (adenocarcinoma, large cell) tumor types,

 and individuals face different prognosis and treatment options depending on the histology. Squamous NSCLC (about 25%-30% of all lung cancers) is typically smoking-related, whereas nonsquamous NSCLC is the most common type of lung cancer in nonsmokers (although smokers are still at greater risk than nonsmokers).

 We examined squamous and nonsquamous NSCLC separately.

We adapted a framework for estimating a therapy’s option value

 to the contexts of metastatic NSCLC and metastatic RCC. In particular, we examined the option value of nivolumab, the first programmed cell death protein-1 (PD-1) inhibitor in both disease states. In order to estimate nivolumab’s option value, a survival curve for patients taking the pre-nivolumab standard of care was needed first. The pre-nivolumab survival curves were modified using mortality hazard ratios (HRs) from clinical trial publications to obtain a nivolumab survival curve in each setting. Forecasting methods were used to project likely future survival improvements patients with either NSCLC or RCC may live to see. The conventional survival gain from nivolumab was obtained by comparing life expectancy with nivolumab, assuming no future innovation, against life expectancy with the pre-nivolumab standard of care. The option value of nivolumab was obtained by comparing life expectancy with nivolumab under 2 scenarios: allowing for likely future survival gains versus assuming no future innovation. Finally, the option value estimates were converted to economic terms. Additional methodological detail is available in the 

Survival trends in metastatic NSCLC and metastatic RCC were estimated using the Surveillance, Epidemiology and End Results (SEER) cancer registry.

 SEER, which tracks cancer incidence and mortality using data reported by registries across the United States, currently covers about 30% of the US population.

Long-term all-cause mortality rates were taken from the Human Mortality Database (HMD).

 The HMD contains detailed population and mortality data for 37 countries, and currently provides US life tables for 1933 through 2013.

Survival curves from SEER were modified to reflect nivolumab patients’ survival using mortality HRs from clinical trial publications.

 Specifically, in both squamous and nonsquamous NSCLC, nivolumab was compared with docetaxel, and in RCC, nivolumab was compared with everolimus. These comparisons reflect those in the trials noted on the nivolumab FDA label.

 The comparator drugs were assumed to represent the pre-nivolumab standard of care. 

We identified 3 study populations: metastatic squamous NSCLC, metastatic nonsquamous NSCLC, and metastatic RCC. First, we identified patients with NSCLC in SEER by requiring a primary cancer site of “lung or bronchus” or “trachea.” We distinguished squamous from nonsquamous histology types using 

International Classification of Diseases for Oncology, Third Edition

) histology codes. For both histology types, we defined metastatic NSCLC as American Joint Committee on Cancer (AJCC) stage IV, IIIb, or III with no surgery. The latter category addressed the fact that not all patients have a substage listed (eg, IIIa vs IIIb), and a lack of surgery suggested the case was inoperable.

To identify the RCC population, we required a primary site of “kidney,” “renal pelvis,” or “ureter,” and identified the RCC histology using 

 histology codes. To limit to patients with RCC with metastatic disease, we selected the “distant” stage using “SEER historic stage A.” Although AJCC staging is commonly used in clinical practice for treatment decisions and in clinical trials for eligibility,

 it was not well documented for RCC in SEER before 2004. Because metastatic RCC is a smaller sample than NSCLC, preserving sample size was important. Moreover, in order to track longer-term trends, it was necessary to be able to identify patients before 2004. In general, SEER distant stage approximates AJCC stage IV. Therefore, we identified metastatic RCC using the SEER distant stage. 

We estimated the option value of nivolumab in the 3 populations by taking these 5 steps.

Step 1: estimate pre-nivolumab survival curves. 

To establish a pre-nivolumab baseline, we estimated survival curves using the SEER data. We identified metastatic squamous and nonsquamous NSCLC and metastatic RCC patients diagnosed between 2001 and 2010. Survival curves for each population were estimated parametrically using a log-normal distribution. Covariates included age at diagnosis, age squared, gender, race, ethnicity, marital status, tumor grade, and a quadratic time trend.

Our survival estimation was limited by the patients’ duration of follow-up in SEER. Although we preferred using patients with a more recent diagnosis due to the pace of innovation in cancer care, this limited the number of years of follow-up available. Rather than project survival for many years beyond the duration of follow-up in the SEER data, we assumed that the minority of patients surviving more than 20 years past diagnosis had the same mortality rates as the general population, using US life tables from the HMD.

Specifically, we calculated mortality rates in the first 20 years after diagnosis using the estimated survival curves from SEER and then appended the HMD mortality rates. This lifetime mortality table was used as an input in the Lee-Carter model described in Step 2. 

The SEER survival curves estimated in Step 1 showed a clear trend of improving survival over time in all 3 populations. Based on these trends, we used the Lee-Carter method

 to forecast survival improvements for each population. The Lee-Carter method involves modeling age-specific mortality rates as a function of a long-term trend in overall mortality improvements and an age-specific response to the overall trend. This method is widely used as a benchmark for long-term mortality forecasts, including by the US Census Bureau and the Social Security Administration.

Step 3: estimate survival with nivolumab.

 To obtain survival curves for patients taking nivolumab in each population, we began with the pre-nivolumab survival curves estimated in Step 1. Based on the FDA label, which calls for second- or later-line use in both the NSCLC and RCC indications,

 we assumed that nivolumab would be given as a second-line therapy starting 1 year after diagnosis. Following the clinical trial evidence,

 we assumed that nivolumab patients would experience diminished mortality risk compared with patients on the prior standard of care. Specifically, we applied the mortality HRs from nivolumab trial publications to the pre-nivolumab survival curves estimated in Step 1. We conservatively assumed that any mortality benefits of nivolumab would disappear 4 years after beginning nivolumab therapy.

Step 4: calculate option value in terms of life expectancy gains. 

Using the survival curves from Steps 1 and 3 and the survival forecasts from Step 2, we calculated the undiscounted life expectancies of patients in each population for several scenarios. First, we calculated life expectancy under the pre-nivolumab standard of care, assuming no future innovation; call this 

. Second, we calculated life expectancy with nivolumab, assuming no future innovation; call this 

. Finally, we calculated life expectancy with nivolumab allowing for future innovation based on the forecasts from Step 2; call this 

We calculated the conventional survival gain with nivolumab by comparing life expectancy with nivolumab to that with the pre-nivolumab standard of care. Specifically:

Conventional survival gain with nivolumab = 

To calculate the option value of nivolumab, we compared life expectancy with nivolumab assuming no further innovation to expected lifetime with nivolumab allowing for future innovation. Specifically:

Step 5: calculate option value in economic terms. 

Finally, we expressed the conventional and option values in economic terms by applying an economic value of a life year to the additional life years gained. We selected a mid-range value from the literature of $150,000 per life year

 and applied a discount rate of 3% to obtain present discounted values. The per-patient option value was multiplied by the estimated size of each patient population to obtain the total economic values in each population. Population sizes were estimated using several sources.

To test the robustness of the study results to the underlying assumptions, several sensitivity analyses were performed. First, we estimated an alternative parametric survival model (log logistic). This tested the sensitivity of the results to our choice of parametric model. Second, we estimated survival exclusively using the parametric model rather than taking HMD mortality rates after 20 years. This tested the effect of our use of the HMD data. Third, we assumed nivolumab would be taken in the first year after diagnosis, rather than the second. This reflected the fact that metastatic cancer can progress rapidly and, therefore, some patients could potentially reach second-line therapy in the first rather than second year from diagnosis. Fourth, we varied the duration post diagnosis that nivolumab would be effective, from 2 to 10 years. This addressed the limited duration of clinical trial data and the accompanying uncertainty about the duration of nivolumab’s effects. Fifth, we estimated nivolumab’s mortality benefits using the confidence intervals of the published mortality HRs in order to address uncertainty about the exact magnitude of nivolumab’s effects of survival. Sixth, we varied the discount rate from 0% to 6%. Seventh, we varied the value of a life year from $50,000 to $250,000.

Option value is the benefit a therapy provides patients by enabling them to survive to the next innovation.

October 2017

Is Higher Patient Satisfaction Associated With Better Stroke Outcomes?

Improving Care Transitions: Complex High-Utilizing Patient Experiences Guide Reform

The Option Value of Innovative Treatments for Non–Small Cell Lung Cancer and Renal Cell Carcinoma

Statewide Data Infrastructure Supports Population Health Management: Diabetes Case Study

A Health Plan's Investigation of Healthy Days and Chronic Conditions

Trends in Hospital–Physician Integration in Medical Oncology

Boosting Workplace Wellness Programs With Financial Incentives

Use of Patient-Reported Outcomes and Satisfaction for Quality Assessments

Low-Value Antibiotic Prescribing and Clinical Factors Influencing Patient Satisfaction

Countywide Physician Organization Learning Collaborative and Changes in Hospitalization Rates