Abstract
The topic of the research is Hospital Readmissions and CMS Penalties. The area has been well explored in the current literature. However, this study aims to reach conclusions from a different perspective. It is imperative to understand ways to reduce readmissions in U.S. hospitals. Secondary data were collected from the available literature on the subject. The inductive approach was adopted to undertake this study. The variables used are the conditions and procedures of patients, age and sex, the type of admission (emergency or not), and the presence of a hospital stay during the previous six months. However, the primary and secondary sources are somewhat similar. The total number of articles extracted from the literature is 30. The detailed analysis of the literature included fact sheets, charts, and graphs showing trends in readmissions in U.S. hospitals.
The results reveal that readmissions in the U.S. have varied over a decade. There are different ways to alter readmission rates in U.S. hospitals. There has been a constant increase in the number of readmissions over time. The risk of potentially preventable readmission was calculated from a Poisson regression model while considering the time of exposure to the risk of readmission. The findings show that most patients were alive, and the risk of readmission was significantly higher among them.
There is a set of recommendations that can improve readmission rates. U.S. hospitals need to focus closely on the issue at hand. Because examining medical records is relatively expensive, it is recommended that analysis target services where readmissions are unusually numerous. If readmission rates are high in all services, it may be useful to examine a random sample of the records of identified readmissions.
Introduction
Background of the Study
There are several reasons for using the readmission rate as a quality indicator in hospitals. First, it is well documented that an early discharge or inadequate treatment during hospitalization can lead to readmission. Second, readmission events are frequent, whereas deaths may occur as a result of a wide range of diseases. Third, the necessary data to calculate readmission rates and adjust them according to cases are routinely available in all United States hospitals. Not all readmissions are problematic. Some are planned at the time of the patient’s discharge, for example, a cholecystectomy after a hospital stay for cholecystitis. Some authors have proposed considering only emergency readmissions that become necessary within a month after discharge. This procedure, however, is problematic. Numerous emergency readmissions are, in fact, due to a new illness unrelated to the previous hospitalization. A patient may, for example, be hospitalized following a car accident three weeks after being discharged from the hospital following a heart attack (Tsai, Joynt, Gawande, & Jha, 2013).
Some planned readmissions are also justified by the treatment of a complication or by reoperation at the same surgical site. Finally, it is necessary to distinguish between planned and expected readmissions. A birth that follows hospitalization for a difficult pregnancy always takes place in an emergency setting (unplanned), although it is clearly expected. The situation is similar in the case of transplants, given that the patient undergoes in-depth evaluations to ensure that he or she is a candidate, but the date of readmission to receive the organ is generally unknown. It is therefore not possible to define whether a readmission is expected based solely on the admission procedure (planned or emergency). For these reasons, some situations require a clinical comparison of the medical data from the first hospitalization with those from the readmission. The analysis of several thousand hospitalizations has allowed the development of a screening algorithm for potentially avoidable readmissions. This algorithm has demonstrated excellent sensitivity (detection of virtually all problematic cases).
Also, a study involving 49 U.S. hospitals allowed the development and approval of an adjustment model that considers patients’ readmission risks according to age, sex, and state of health. As a result of these scientific studies, the National Association for the Development of Quality in Hospitals and Clinics (ANQ) has decided to include potentially avoidable readmission rates among the indicators used to monitor quality in U.S. hospitals. The ANQ considers this index useful because it shows differences between hospitals by comparing observed and expected rates based on patient characteristics. The use of a secure algorithm based entirely on existing medical statistics also avoids the costly collection of additional data (Joynt & Jha, 2012). Hospital medical statistics in the U.S. have the added advantage of being able to identify readmissions occurring in other hospitals through the confidential code introduced ten years ago by the Federal Statistical Office. In this way, a hospital whose patients prefer another institution in the event of a complication is treated in the same way as a hospital that retains the patient’s confidence at readmission (Berenson, Paulus, & Kalman, 2012).
Overview of the Study
Hospital readmissions are very common among patients with cardiovascular diseases in the United States. Various socioeconomic factors have an impact on these readmissions among people 65 years or older (Damiani et al., 2015). The Affordable Care Act added section 1886(q) through Section 3025 of the Social Security Act to establish the Hospital Readmissions Reduction Program (HRRP). The HRRP requires the Centers for Medicare and Medicaid Services (CMS) to reduce reimbursements to inpatient prospective payment system (IPPS) hospitals with excess readmissions for discharges beginning on October 1, 2012 (Centers for Medicare & Medicaid Services [CMS], 2016). The principles that apply to this program are set out in subpart I of 42 CFR part 412 (§412.150 through §412.154).
Nowadays, hospital readmissions receive particular attention because of the importance of reducing poor-quality care and the excessive spending they represent. A 2009 study mentions that close to 20% of Medicare patients are readmitted to the hospital within a 30-day window after discharge, representing up to $17 billion in excess annual costs to the government (Berenson et al., 2012). The most important factors in avoiding rehospitalization include hospital-acquired infections and other complications, premature discharge, failure to coordinate and reconcile medications, poor communication among hospital personnel, patients, caregivers, and community-based clinicians, and poor planning for care transitions (Berenson et al., 2012). Policymakers believe that reducing readmission rates would improve patient care and reduce medical expenses (Joynt & Jha, 2012).
Although some studies agree that hospital readmissions are an important topic for hospitals, they also believe that policymakers’ emphasis on 30-day readmissions is mistaken for three reasons. The first reason is that the metric used by hospitals is problematic. Only a small proportion of hospital readmissions within 30 days of the first discharge may be preventable, and there are also patient and community factors outside hospitals’ control that lead patients back to hospitalization. Moreover, it is not clear that readmissions usually represent poor hospital quality. These readmissions could be attributable to low mortality rates or greater accessibility to hospital care (Joynt & Jha, 2012). The second reason identified by these authors is that other policies may need more attention than readmissions. The final reason is that because hospitals are paying substantial attention to readmissions, they sometimes minimize the importance of other issues related to improving healthcare, such as patient safety (Joynt & Jha, 2012).
What should be more important to policymakers, health professionals, and the public in the healthcare industry is providing and receiving better healthcare for the nation at a lower cost.
Research Question
- Highlight the factors of hospital readmissions and CMS penalties.
Literature Review
Re-Hospitalization in the Medicare Population
Re-hospitalization in the Medicare population may be associated with different factors, including failure to understand or follow physician instructions, recurrence of disease, or lack of follow-up care. Moreover, enrollment in Medicare is expected to grow, and consequently these readmissions could increase hospital expenses (Miletic et al., & Kaye, 2014). These authors conducted a study to determine whether a telephone intervention or support call reduces rehospitalization compared with the corresponding control group. The study analyzed patient data after telephone support was provided following hospital discharge (Costantino et al., 2013).
The intervention was received by 48,538 Medicare enrollees, and 4,504 (9.3%) of this sample were rehospitalized within 30 days of discharge, compared with 5,598 controls (11.5%, P < 0.0001) (Costantino et al., 2013). The authors found a direct correlation between the timing of the telephone intervention and the rate of re-hospitalization. Consequently, the closer the patient receives the phone call to the date of initial discharge, the greater the reduction in the number of rehospitalizations. Additionally, in the intervention group, visits to the emergency room were reduced compared with controls (8.1% vs. 9.4%, P < 0.0001). Visits to the physician’s office increased (76.5% vs. 72.3%, P < 0.0001); this suggests that the intervention motivated enrollees to seek advice to avoid readmission. Overall, the total savings were $499,458 for the enrollees who received the intervention, with total savings of $13,964,773 to the healthcare insurance plan (Costantino et al., 2013). Assisting patients after discharge can help reduce hospital readmissions and expenses.
The main issue concerning readmissions is that the number of Medicare members is increasing each year, and it is crucial for Medicare to try to control healthcare expenses. It is essential to seek to cut these costs by avoiding readmissions after discharge, especially those that can be prevented (Costantino et al., 2013). In 2005, Medicare readmissions reported by the U.S. government resulted in expenditures of $12 billion, and in 2004, the amount spent on preventable readmissions was $17.4 billion (Costantino et al., 2013). The current estimated cost of re-hospitalizations within 30 days of discharge is reported to be $44 billion a year, including other patients’ total healthcare costs.
It is considered that many hospital readmissions can be avoided because they may represent poor-quality care or poor transitional care (Costantino et al., 2013). These authors also think that readmissions are due to different factors such as early discharge because of the need for hospital beds, lack of follow-up care for discharged patients, misunderstanding of discharge instructions, and a shortage of help for patients who need continuing care at home (Crockett, 2017). This program provides telephone support for patients after discharge, making sure they follow the physician’s instructions and receive enough support from their families. It could help reduce hospital readmission rates. However, future studies should evaluate transitional-care costs and utilization (Costantino et al., 2013).
Impact of Social Factors
One of the most important topics considered in many studies is the impact of social factors after a patient’s discharge. These variables affect readmissions within the 30-day period because they are outside the hospital setting and cannot be controlled by the hospital. The social factors evaluated in the majority of the CAP studies include race, age, and gender. This study mentions that worse outcomes are associated with older age, that the results for gender were mixed, and that non-White patients were the racial group most affected (Faverio et al., 2014).
For heart failure, older age was significantly related to higher patient-readmission rates, but the variable of gender showed mixed results. Non-White patients were also the racial group most affected, although they had lower mortality (King et al., 2013; Calvillo-King et al., 2013).
Hospitals that receive a higher proportion of low-income patients have a greater likelihood of being penalized by the HRRP program (American Hospital Association, 2015).
(American Hospital Association, 2015).
Fiscal Years
In the Fiscal Year (FY) 2012 IPPS final rule, CMS stated its policies for hospital readmissions under the HRRP, including the definition of readmission. Readmission is admission to a hospital within 30 days of discharge (CMS, 2016). CMS adopted re-hospitalization measures for acute myocardial infarction (AMI), heart failure (HF), and pneumonia (PN). CMS also established a method to assess the excess readmission ratio for each applicable condition, which is used in part to calculate the readmission payment adjustment.
The excess readmission ratio is a measure of hospital performance compared with the national average for hospitals treating patients with a consistent set of conditions (CMS, 2016). CMS established a risk-adjustment methodology approved by the National Quality Forum (NQF). The excess readmission-ratio system includes adjustments for certain patient characteristics, including demographics and comorbidities.
Additionally, the addition of new conditions to the HRRP program in Fiscal Years 2013-2015 increased the percentage of hospitals penalized in the United States.
(American Hospital Association, 2015)
These hospitals experienced an increase in the number of conditions included in the readmission program, such as total hip and knee replacement.
Hospital Readmissions Reduction Program (HRRP)
The Hospital Readmissions Reduction Program (HRRP) is one of the payment initiatives implemented after the enactment of the Patient Protection and Affordable Care Act (PPACA). This program penalizes hospitals through a Medicare reimbursement-rate reduction of 1% when they have excessive patient re-hospitalization in the first year for conditions including heart attack, pneumonia, or congestive heart failure (Costantino et al., 2013). If hospital readmissions do not improve by the third year, an additional 2% reduction will be added. Another study also mentions that in the United States, among patients enrolled in Medicare fee-for-service who were discharged from a hospital, 19.6% were readmitted within 30 days (Damiani et al., 2015).
Another study mentions the addition of chronic obstructive pulmonary disease (COPD) to the HRRP by the U.S. CMS in October 2014, as well as the inclusion of acute exacerbation of chronic obstructive pulmonary disease (AECOPD) (Feemster & Au, 2014). The objective is to encourage hospital quality and decrease expenses by creating financial incentives for hospitals to avoid readmissions (Feemster & Au, 2014). These authors also mention that more than 2,000 hospitals in the U.S. had been fined by that date. This resulted in approximately $280 million in penalties for fiscal year 2013 (Feemster & Au, 2014).
CMS expanded the conditions to include all-cause unplanned rehospitalizations for AECOPD in October 2014. COPD is an important disease to measure for readmissions because, in the U.S., it represents more than 1.5 million visits to the emergency room and 725,000 hospital admissions. These events represent annual healthcare expenses of up to $60 billion in the U.S. (Feemster & Au, 2014). About 34 to 40% of hospitalized AECOPD patients did not receive recommended treatments, while approximately half of the patients received at least one incorrect or potentially dangerous treatment (Feemster & Au, 2014).
The proportion of patients with AECOPD who are rehospitalized within 30 days after discharge is nearly 22.6% for all causes, demonstrating the importance of the issue to patients, hospitals, and payers (Feemster & Au, 2014). Hospitals measure readmissions by assessing administrative and claims data. Social factors showed some influence on AECOPD readmission rates, which were 22% higher among patients living in low-socioeconomic-status areas and higher among Black patients than among other racial or ethnic groups (Feemster & Au, 2014). In fiscal year 2012, more than 3,000 hospitals were subjected to high penalties. A noticeable proportion of COPD patients have low-income status. CMS should try to provide additional support and funds to hospitals to help patients transition from the hospital to their homes through the Community-based Care Transition Program (Berenson et al., 2012). Approximately 33% of 1,904,640 deaths among people aged 65 years and older in the U.S. occurred in a hospital in 2013, although most of these Medicare enrollees would have preferred to die at home (Gorina et al., 2015). The mortality rate among patients hospitalized for pneumonia was 12.1% within the first 30 days of hospitalization (Gorina et al., 2015).
Close to 1 in 5 patients enrolled in Medicare fee-for-service who were discharged from a healthcare institution were readmitted within 30 days (Hansen, Young, Hinami, Leung, & Williams, 2011). One study compares three different data-collection methods used to evaluate unplanned readmissions among patients who had colorectal surgery from July 2009 to November 2011. The methods used were the NSQIP clinical-reviewer method, physician medical-record review, and the University Health System Consortium (UHC) administrative-billing-data method.
The researchers found 735 patients who had colorectal surgery and were evaluated for hospital readmission. The NSQIP method identified 14.6% (107 readmitted patients), whereas UHC identified 17.6% (129 readmitted patients). For readmissions associated with the index admission, NSQIP detected 72%, while physician chart review detected 83%. The UHC method found that 51% of rehospitalizations were associated with the index admission, while physician chart review found 86%. It is important to note the significant discrepancy among the methods. According to the UHC method, 66 of the 129 readmissions (51%) were considered preventable. In contrast, physician chart review found that 112 of 129 rehospitalizations (87%) were necessary when the patient returned to the hospital.
These methods showed that most readmissions were attributable to surgical-site infections (46 of 129, or 36%) and dehydration (30 of 129, or 23%). Of the 129 patients, 41 (31.8%) who presented complications but received good care after discharge did not need readmission. This study suggests that results on readmission rates also depend on the type of method used to evaluate readmissions. Sometimes the methods produce different outcomes, and these results can affect hospital penalties.
This type of study reflects the need for hospitals or healthcare organizations to find better ways to collect data in order to have accurate information about patient readmissions and demonstrate to CMS the correct number at the time of an audit. Keeping all this information in a file, archive, software application, or program will help identify genuine readmissions and avoid inappropriate penalties (Berenson, 2012).
The CMS determined projected readmission rates for hospitals in the U.S. These rates covered pneumonia, acute myocardial infarction, and congestive heart failure from July 2008 to June 2011 (Joynt & Jha, 2013a). CMS also adjusted these readmission rates for sex, age, and coexisting conditions. CMS examined the expected readmission rates for each hospital, and hospitals that exceeded these rates were penalized.
CMS reported that close to 66% of hospitals in the U.S. received penalties of up to 1% of their reimbursement for Medicare members (Joynt & Jha, 2013a). By 2015, CMS was expected to expand the penalties to 3%, and by 2013 CMS was predicted to recover $280 million from the 2,217 hospitals that were penalized.
There is considerable controversy regarding hospital penalties for high readmission rates. There are two main reasons why this policy is debatable. The first is whether hospitals should be held responsible for rehospitalizations after patients are discharged with physician instructions and medication, when the continuation of care depends substantially on the patients. The second point of contention concerns the different types of measures used to evaluate hospital readmission rates (Burke & Coleman, 2013).
Two groups of patients are more susceptible to hospital readmissions: those who have a severe illness and those who lack post-discharge care because of low socioeconomic status (Joynt & Jha, 2013a). Some readmission measures may be unfair because they do not fully consider the complexity of each patient’s disease or disability. Instead, socioeconomic factors may create a significant disadvantage for hospitals that care for large numbers of patients with severe illnesses and low socioeconomic status and that are therefore at risk of substantial penalties.
There are circumstances in which hospitals may be able to influence the reported results. First, the HRRP has encouraged investment in activities and programs to prevent hospital readmissions, and hospital readmissions decreased from 15.5% in 2009 to 15.3% in 2011 (Joynt & Jha, 2013a). It would be beneficial if these reductions reflected genuine improvements rather than hospitals shifting patients to emergency rooms or observation areas.
This decrease was greater than CMS expected. It is therefore important to emphasize that there is evidence that safety-net hospitals, as well as teaching hospitals, are highly penalized by the HRRP (Joynt & Jha, 2013a).
(Joynt & Jha, 2013a).
The figure shows the proportion of hospitals subject to the readmission penalty (Panel A) and the average amount of the penalty (Panel B), according to the proportion of hospital patients receiving Supplemental Security Income. Medicare Payment Advisory Commission data are presented.
Considering these two perspectives, other actions may be worthwhile to maintain the gains that have been achieved while preventing excessive penalties for hospitals that serve underserved patients. First, readmission rates need to be adjusted for socioeconomic factors so that safety-net and non-safety-net hospitals can be compared more fairly. This would help prevent safety-net hospitals from being penalized simply because they treat a larger number of severely ill patients.
Second, it would be important for the HRRP to implement penalties according to the timing of readmission. Readmissions that occur during the first days after discharge might indicate that a patient had inadequate post-discharge care or did not follow instructions appropriately, whereas readmissions occurring four weeks after the initial discharge may reflect the severity of the patient’s illness (Joynt & Jha, 2013a). Moreover, a rehospitalization occurring three hours or three days after discharge could be weighted more heavily by the algorithm than a 30-day readmission.
For this reason, hospitals that care for a significant number of patients with socioeconomic problems and severe diseases would benefit from coordinating and planning post-discharge care to avoid short-term rehospitalizations and prevent penalties associated with these patient factors. Lastly, giving incentives to hospitals for low mortality rates could compensate for the risks that hospitals face from readmissions. This means that low hospital mortality rates would be considered in readmission metrics instead of penalizing hospitals for keeping patients with severe diseases alive even when some are readmitted (Joynt & Jha, 2013a). Under the current CMS approach, large teaching hospitals may be more at risk of being penalized for high readmission rates despite low mortality rates, whereas hospitals with low readmission rates may perform better in the metric even if they have higher mortality rates. These authors suggest that one technique for combining these two outcomes is to evaluate patients for 30 days after discharge while considering the number of days they are alive and out of the hospital (Joynt & Jha, 2013a).
Similarly, Joynt and Jha found in another study that large teaching hospitals and safety-net hospitals (SNHs) are more vulnerable to being penalized by the HRRP (Joynt & Jha, 2013b).
(Joynt & Jha, 2013b).
Policies should change and adapt to the needs of the public and organizations. HRRP penalties for hospitals that care for the most vulnerable patients and those with socioeconomic disadvantages could be reduced if the HRRP changed its metrics, provided incentives, or modified guidelines to help hospitals serving the poorest patients in the U.S. Such changes could support programs that improve coordination of care after patients are discharged and during their recovery at home. These events occur outside hospital walls, but with a proper coordination plan for discharged patients, hospitals would be less affected by HRRP penalties.
Methodology
Research Design
The study uses an inductive approach because there has been substantial research on hospital readmissions and how they can be reduced. However, this research focuses specifically on hospital readmissions and CMS penalties. An inductive study builds on observations and findings from previous research while exploring them in a new direction. After recovery from a serious illness, some patients are susceptible to new complications, many of which require readmission to the intensive care unit (ICU). This is associated with increased mortality and longer hospital stays. Early identification of patients at risk of ICU readmission can facilitate the appropriate allocation of resources to prevent morbidity and mortality.
Research Approach
A qualitative research approach will be used to conduct the study. Qualitative research helps researchers focus on human experiences and examine factors such as beliefs, customs, attitudes, and emotions (Tsai et al., 2013).
Data Collection
Data will be collected from secondary sources such as the internet, published papers, articles, and online libraries. Thirty articles have been selected from 900 studies because these materials fulfill the inclusion criteria required for the research.
| CRITERIA | INCLUSION | EXCLUSION |
| From 900 articles | 800 articles were excluded because of their titles and abstracts. | |
| 50 articles | Were excluded after more detailed reading. | |
| 40 articles | Were selected for full-text review. | |
| 30 articles | Were selected because they met the inclusion criteria. |
Data Analysis
The search was conducted in electronic databases for the period from January 1, 2006, to August 22, 2016. Studies associated with reduced readmissions, socioeconomic factors, elderly patients, and CMS penalties in hospitals in the United States were included.
Sources of Information
Google Scholar and eBook libraries are the primary sources of data. The keywords used were elderly, readmissions, heart failure, social factors, Medicare, Medicaid, and penalties.
Results
A readmission is considered potentially avoidable if it was not scheduled at discharge from the previous hospitalization, if it is caused by at least one condition already known at the time of discharge, and if the readmission occurs within thirty days. Readmissions related to transplants, childbirth, chemotherapy, radiotherapy, or surgical intervention following a hospital stay for medical appointments are examples considered expected. Readmissions for a new condition that was not present at the time of the previous hospitalization are considered unavoidable. The term “potentially preventable” means that, under ideal circumstances, the readmission would not have been expected at the time of discharge. Thus, it is an undesired event whose causes can be manifold. The algorithm identifies unwanted readmissions, as demonstrated by the sensitivity (96%) and specificity (96%) of identification, but this does not mean that all readmissions can be avoided. That is why excessively frequent readmissions should be analyzed critically to identify their causes. The 30-day period is commonly indicated in the scientific literature and has been confirmed by studies conducted on U.S. data.
Determination of the Population at Risk
The population at risk of potential readmission includes all patients who were hospitalized, discharged alive, not transferred to another hospital, and were U.S. residents. Analysis of the data showed that the U.S. measurement of readmission rates could be altered by the inclusion or exclusion of hospitalizations for surgeries that could also be performed on an outpatient basis. Because these interventions are frequent in some hospitals and rare in others (from 4% to 30% of planned surgery), they were excluded from the calculation. For the same reason, hospitalizations for sleep apnea were also excluded. Patients residing abroad are excluded because they may be readmitted in another country, which would distort comparisons between hospitals. Infants are excluded because the readmission indicator is intended to assess the quality of preparation for the discharge of sick patients. Measurement of the readmission rate takes into account the time elapsed after a patient’s discharge. As a result, a patient readmitted to the same or another hospital is no longer at risk of the initial readmission, resulting in censoring of the exposure time at risk. Each readmission may, however, be followed by another readmission.
Calculation of the Adjusted Readmission Rate
The risk of potentially preventable readmission is calculated from a Poisson regression model that considers the time of exposure to readmission risk. The variables used are patients’ conditions and procedures, age and sex, the type of admission (emergency or not), and the presence of a hospital stay during the previous six months.
To calculate expected readmission rates as a function of case mix in 262 hospitals, more than 3.2 million discharges from U.S. hospitals between December 1, 2003, and November 30, 2007, were used (Tsai et al., 2013). Hospital data whose coding quality was questionable were discarded. The clinical categories in question originate from the SQLape® classification system, which considers all conditions and procedures presented by patients, regardless of their rank (primary or secondary). SQLape is software used by hospitals for patient-quality and cost management. Some categories are associated with low risk (obstetrics, ENT infections, and skin diseases), while others are associated with high operative risk (amputations, transplants, coronary bypass, major interventions involving the digestive tract, etc.). High-risk diagnostic categories include mainly chronic, recurrent, or malignant diseases: tumors, agranulocytosis, ischemic disease, cirrhosis, respiratory failure, and mental illness (depression, schizophrenia, drug addiction, and anorexia nervosa) (Merkow et al., 2015). A patient who has malignant cancer, anorexia, or chronic renal insufficiency may have a risk around ten times higher. These conditions are often comorbid and unrelated to the main diagnosis, which explains why a DRG grouper cannot be used to determine this type of risk. Even comorbidity indexes (for example, Charlson) were considered poorly predictive because they contain few diagnostic categories and no operative criterion. Because the predictive model presents statistical uncertainty, a 95% confidence interval was used to define minimum and maximum adjusted rates (Nedza et al., 2016).
On a practical level, the adjusted readmission rate was calculated by multiplying the average rate observed in Switzerland by the ratio between the observed and expected rates of each hospital. The following example illustrates the procedure (Rau, 2014). Overall rate observed in Switzerland: 5%. Rate observed at Hospital H: 6%. Expected rate at Hospital H: 4%. Adjusted hospital readmission rate H: 7.5% (= 5% × 6% / 4%). This procedure has the advantage of considering each hospital’s patient population when judging performance and allows hospitals to be compared using the adjusted rate. In this example, the adjusted rate is higher than the observed rate, reflecting the fact that patients admitted to the hospital in question had a lower risk of being readmitted than those at other hospitals (Shah et al., 2014).
Discussion
Ideally, an indicator must meet several requirements: utility, accuracy, absence of distortion, relevance, precision, reliability and reproducibility, reasonable cost, comparability, and availability. Reducing the number of potentially avoidable readmissions is useful for lowering costs and improving patient safety. The indicator’s accuracy is supported by excellent sensitivity and specificity in identifying relevant cases (numerator) and by an appropriate definition of the population at risk (denominator). Distortion is reduced by excluding hospitalizations for surgical interventions that could also be performed on an outpatient basis and by including readmissions to other hospitals. Differences between hospitals in observed and expected rates demonstrate the usefulness of the indicator. Confidence intervals calculated from SQLape® are sufficiently narrow to highlight significant differences between hospitals. Coding quality is examined with the instrument to detect reliability problems. The tool is based on data routinely available in all hospitals, which helps limit the cost of producing the indicator (Swinburne et al., 2017).
The calculation of expected readmission rates considers all available information on patients’ health status to ensure comparability between hospitals. To avoid misinterpretation, users should be aware of two limitations of the system. The first is the waiting time required to obtain results, given that absolute values may become available only after more than a year. The calculation takes into account readmissions to other hospitals. This means that the data collected by the Federal Statistical Office must be complete and validated. An interim rate can be determined by a hospital by installing the tool internally, but the external-readmission rate must be estimated using observations from the previous year. The second limitation is the difficulty of documenting the causes of readmission. About a quarter of potentially avoidable readmissions can be attributed to problems for which hospitals may bear responsibility, for example, surgical complications, medication side effects, and premature discharge. Half of readmissions are tied to difficulties in managing the situation on an outpatient basis. These may involve a shortage of care after hospitalization, inappropriate patient behavior, or aggravation of disease that, in some cases, could have been avoided with better organization of outpatient follow-up care (Tsai et al., 2013).
Finally, another quarter of readmissions are caused by the spontaneous evolution of disease; in these cases, it is not possible to identify errors in the care provided. It should be emphasized that the expected rate takes these situations into account, and it cannot be expected that a hospital will have no potentially avoidable readmissions. The value of the instrument is that it identifies suspicious readmissions without forcing hospitals to review every case. A low adjusted rate is reassuring, and detailed analysis can then focus on hospitals or services with excessively high levels.
The measurement of potentially avoidable readmission rates can be distorted if the quality of hospital medical statistics is insufficient. Data-quality requirements concern completeness, accuracy, and compliance with the coding of diagnoses, procedures, and administrative data (mode of admission, days of operation, etc.). If data quality appears suspicious, for example, if patient numbers vary unexpectedly from one year to another, a warning is issued. Data quality is evaluated in Excel tables, which allows the format of data provided to SQLape® to be checked for consistency and corrected if necessary. If the data provided are identical, the results should also be identical.
Analyses carried out in Switzerland have demonstrated so far that almost 50% of readmissions were linked to relapse or aggravation of a disease already present at the time of admission. These disorders were often comorbidities that had not justified the previous hospitalization. Typically, such readmissions cannot be attributed to only one hospital, which may have had only an indirect influence on outpatient care. Nevertheless, it is still worthwhile to emphasize the hospital’s partial responsibility. Subsequent outpatient care is often provided by hospital doctors (e.g., surgeons and oncologists) or hospital polyclinics. On the other hand, the hospital is obliged to arrange subsequent outpatient treatment, for example, by making the first appointment with the doctor, sending necessary information in a timely manner, and providing services immediately to the requested address.
Experience shows that hospital nursing staff have little contact with home-care services, which often take over late, particularly when patients are discharged over the weekend. Sometimes a readmission occurs because the patient did not follow or understand the recommendations; even in these cases, the hospital can help improve the situation by providing information to the patient or family. A quarter of potentially avoidable readmissions is due to the progression of disease after discharge despite optimal treatment. Unfortunately, it is not possible to isolate these cases using medical-statistics data, so they remain a confounding factor. It should, however, be remembered that expected rates take patient health into account, and hospitals therefore should not be penalized unfairly.
Analysis of the causes of readmissions is possible if the patient returns to the same hospital, but it is obviously more complicated when he or she is admitted to another hospital. Respect for confidentiality prevents the second hospital from communicating the patient’s name to the first hospital. Requesting permission from patients at the time of discharge is not an adequate solution because patients have the right to keep the reasons for their dissatisfaction private if they prefer to be readmitted elsewhere. If a hospital is confronted with a high rate of readmissions of its patients to other hospitals and an overall rate that is too high, it is advisable to rely on an external medical auditor who will not reveal the names of readmitted patients. In this case, a solution might be to assign a unique patient number to the hospitals in question.
CMS’ Hospital Readmission Penalties Worked Under Affordable Care Act
Medicare began using a new tactic to get hospitals to reduce costly, unnecessary readmissions. The Hospital Readmissions Reduction Program, established in the 2010 Affordable Care Act, allows CMS to withhold inpatient prospective payments from short-term acute-care hospitals with excessive readmissions for certain conditions. Since then, debate has continued over whether the program is effective at improving hospitals’ quality of care (Fontanarosa & McNutt, 2013).
Opponents cite concerns about unintended consequences and severe penalties for institutions that serve sicker patients, while the government maintains that its efforts have borne fruit. A new study, published in the Annals of Internal Medicine, concluded that hospital readmissions across the U.S. declined after the ACA was introduced and that hospitals with the highest readmission rates before 2010 improved the most in the years that followed.
There has been gradual, continuous, and increasing attention devoted to the issue of readmissions (Herrin et al., 2014). Although it is not possible to disentangle the effects of financial penalties, public reporting, and increased attention to readmissions and determine their specific contributions to reduced readmissions, the study showed that financially incentivizing reduced readmissions is associated with lower readmission rates (Kangovi & Grande, 2011).
Figure 1 Percentage Difference In Re-Admission Rates Between High And Low Communities (Kansagara et al., 2011)
The main finding was that a large reduction in rates followed the new law, particularly among lower-performing hospitals. The researchers examined the records of more than 15 million fee-for-service Medicare patients discharged from 2,868 acute-care hospitals for heart attacks, congestive heart failure, or pneumonia from 2000 to 2013. They calculated risk-standardized readmission rates and categorized hospitals into four performance categories: highest, average, low, and lowest. Although overall readmissions fell after 2010—the post-law period—lower-performing groups made greater gains. Risk-standardized readmissions dropped by 69 and 74.5 per 10,000 discharges annually among the highest- and average-performing groups (Kansagara et al., 2011).
Figure 2: Lowest Rate Of Readmissions In The U.S (Kripalani, Theobald, Anctil, & Vasilevskis, 2013).
Among the low- and lowest-performing groups, they dropped by 83.2 and 92.4 per 10,000 discharges, respectively. In other words, the hospitals that were financially penalized the most improved the most. Although readmission penalties did not take effect until October 2012, the researchers counted the first quarter of 2010, when President Barack Obama signed the ACA, as the intervention period (Kripalani et al., 2013).
The fact that hospitals became aware of looming penalties in 2010 and began redesigning care in anticipation of them justified that decision, the authors wrote in the study. These findings come on the heels of an allowance granted to safety-net hospitals in a section of the 21st Century Cures Act, signed into law in mid-December, which adjusts the risk of hospital readmission penalties based on patient mix. Now, when fines are calculated, hospitals that treat higher proportions of poorer, sicker patients will be compared with similar institutions rather than with those that treat healthier, more affluent patients (Krumholz et al., 2013). Stephen Soumerai, who teaches population medicine at Harvard Medical School and was not involved in and had not seen the study, cautioned against extrapolating from it that health outcomes had improved as a result of reduced readmissions.
Conclusion
In some cases, the cause of readmission can be deduced solely from medical statistics, where appropriate with computer assistance. In other situations, analysis of the discharge or readmission letter is necessary because it usually explains why the patient was hospitalized again. Finally, from a medical perspective, it is sometimes necessary to determine whether outpatient treatment was appropriate and whether better-informed outpatient doctors could have helped avoid readmission. Experience shows that the causes of readmission vary when hospitals have normal rates. In these cases, it is not always easy to take measures for improvement. Conversely, when rates are too high, the reasons for readmissions often concentrate in a small number of cases that can be reviewed. The purpose is not to achieve zero potentially avoidable readmissions. That would require a substantial increase in resources to ensure discharge under the best possible conditions, which would generate considerable costs and could be detrimental to other aspects of quality (Joynt & Jha, 2012).
Because examining medical records is relatively expensive, it is recommended that analysis target services where readmissions are too numerous. If readmission rates are high in all services, it may be useful to examine a random sample of records from identified readmissions. If an excessive number of early discharges is suspected, it may be useful to determine whether readmissions are associated with shorter stays, adjusted according to the severity of the case. If readmission rates are consistently too high, it is advisable to monitor them quarterly and ask responsible doctors to systematically document the causes of identified readmissions.
Because the patients have been seen recently, doctors can quickly provide their interpretation of the reason for readmission without first conducting lengthy reviews of the records. The data provided in SQLape® must include data from June 1 of the previous year through 30 days after the last analyzed hospital discharge.
Observed and expected rates can be calculated for distinct periods from detailed hospitalization files. Practical recommendations depend on the results. If the observed rate is below the expected rate, the team caring for patients should be congratulated. A review of clinical records is always possible to understand how the tool works, but it is unlikely to provide much information about improving effectiveness. If the observed rate is higher than the expected rate but lower than the maximum expected rate, the results should be analyzed for each service to isolate those with excessively high levels. These services could benefit from a review of medical records. If the observed rate is higher than the maximum expected rate for the first time, a review of medical records should be conducted. It may be useful to exclude services with average rates and test a random sample; if the issue concerns the hospital overall, the causes of readmissions should be analyzed to determine whether particular services or diseases are involved and whether standard measures across several services would reduce the rates (Chakraborty et al., 2011).
When observed rates remain above the maximum expected rate, quarterly analyses should be performed as soon as medical statistics are complete, and the physician responsible should receive data on his or her patients and be asked to assign the most likely cause of each readmission. This procedure requires organizational effort, but it can be faster than reviewing historical medical records. It should be remembered that standard rates are calculated regardless of whether readmissions take place at the same hospital. It is therefore necessary to correct the observed gross rate based on the data before comparing it with the expected rate. The software used by the hospital for internal analysis is the same and should therefore provide the same results for internal observed rates and individual cases. If the calculation of readmission rates is correct, it is essential that the number of patients be consistent regardless of the year in question. The externally observed rate is higher because it takes into account all U.S. hospitals in which patients may have been readmitted. Even the observed rate may be slightly higher because it considers admissions during the previous six months in all U.S. hospitals (Sosunov et al., 2016).
Recommendations
There are several possible measures to reduce the number of potentially avoidable readmissions if their causes are known. According to the scientific literature, the side effects of drugs represent three-quarters of the adverse events that occur in the month following discharge (Chakraborty et al., 2016). Readmissions caused by these events are rare, but half of the cases are due to medication errors, for example, drug interactions and inadequate monitoring of anticoagulant treatment. Some preventive measures for high-risk patients (those taking multiple drugs, antibiotics, glucocorticoids, anticoagulants, antiepileptics, and hypoglycemic agents) have proven useful (Chakraborty et al., 2016). Surgical complications can also be too frequent for a variety of reasons, including questionable surgical indications or techniques, poor infection-prevention measures, or insufficient team skills or training. An analysis of the profile of readmitted patients could help ensure that the procedures performed correspond to the hospital’s mandate. Some readmissions of this kind are expected, but they should remain below the expected rate after taking patients’ diseases into account.
There may also be other complications, such as thrombosis or embolisms. If they are too numerous, it may be necessary to determine whether adequate preventive measures are being taken. The discharge procedure is certainly a critical transition point (Chakraborty et al., 2016). Communication between hospitals and services caring for patients is often insufficient. Several studies have demonstrated the frequent omission of information relevant to subsequent patient care, including recent test results and follow-up plans. It is established that planning outpatient appointments decreases the risk of readmission (Chakraborty et al., 2016).
Some authors have proposed a checklist for the discharge procedure, but its effectiveness has not been studied. So far, U.S. hospitals have found few early discharges, but if these situations become frequent, primary-care doctors should be warned and more involved in decisions concerning discharge and its organization.
In general, good cooperation among hospital physicians, their outpatient colleagues, nursing staff, patients, and family members helps ensure that discharge is well prepared. However, this process can be overlooked, for example, if the hospital is overloaded (with a consistently very high occupancy rate) or if doctors are regularly under pressure from emergency-room admissions. In the latter case, especially if patients are elderly and suffer from multiple diseases, it may be appropriate to provide a unit responsible for managing discharge (Chakraborty et al., 2016).
References
American Hospital Association. (2015). Trendwatch. Rethinking the Hospital Readmissions
Berenson, R. A., Paulus, R. A., & Kalman, N. S. (2012). Medicare’s Readmissions-Reduction Program – A Positive Alternative. The New England Journal of Medicine. Retrieved from
https://www.nejm.org/doi/full/10.1056/NEJMp1201268#t=article
Burke, R. E., & Coleman, E. A. (2013). Interventions to decrease hospital readmissions keys for cost-effectiveness. The Journal of the American Medical Association, 173(8), 695-698, doi:10.1001/jamainternmed.2013.171
Calvillo-King, L., Arnold, D., Eubank, K. J., Lo. M., Yunyongying, P., Stieglitz, H., & Halm, E. (2013). The impact of social factors on risk of readmission or mortality in pneumonia and heart failure: Systematic review. Journal of General Internal Medicine, 28(2), 269-282. doi:10.1007/s11606-012-2235-x
Centers for Medicare & Medicaid Services [CMS]. (2016) Readmissions Reduction Program (HRRP). Retrieved from https://www.cms.gov/medicare/medicare-fee-for-service-payment/acuteinpatientpps/readmissions-reduction-program.html
Chakraborty, A., Selby, D., Gardiner, K., Myers, J., Moravan, V., & Wright, F. (2011). Malignant bowel obstruction: the natural history of a heterogeneous patient population followed prospectively over two years. Journal of Pain and Symptom Management, 41(2), 412-420. doi: 10.1016/j.jpainsymman.2010.05.007.
Chakraborty, H., Axon, R. N., Brittingham, J., Lyons, G. R., Cole, L., & Turley, C. B. (2016). Differences in hospital readmission risk across all payer groups in South Carolina. Health Services Research. Retrieved from https://www.researchgate.net/publication/ 308738412_Differences_in_Hospital_Readmission_Risk_across_All_Payer_Groups_in_South_Carolina. doi: 10.1111/1475-6773.12579
Constantino, M. E., Frey, B., Hall, B., & Painter, P. (2013). The Influence of a Postdischarge Intervention on Reducing Hospital Readmissions in a Medicare Population. Population Health Management, 16(5), 310-316. doi:10.1089/pop.2012.0084
Damiani, G., Salvatori, E., Silvestrini, G., Ivanova, I., Bojovic, L., Iodice, L., & Ricciardi, W. (2015). Influence of socioeconomic factors on hospital readmissions for heart failure and acute myocardial infarction in patients 65 years and older: Evidence from a systematic review. Dove Medical Press, 10, 237-245. doi:10.2147/CIA.S71165
Faverio, P., Aliberti, S., Bellelli, G., Suigo, G., Lonni, S., Pesci, A., & Restrepo, M. I. (2014). The management of community-acquired pneumonia in the elderly. European Journal of Internal Medicine, 25(4), 312-319. doi: 10.1016/j.ejim.2013.12.001
Feemster, L. & Au, D. H. (2014). Penalizing hospitals for chronic obstructive pulmonary disease readmissions. American Journal of Respiratory and Critical Care Medicine, 189(6), 634-639. doi:10.1164/rccm.201308-1541PP
Fontanarosa, P., & McNutt, R. A. (2013). Revisiting hospital readmissions. The Journal of the American Medical Association, 309(4), 398-400. doi:10.1001/jama.2013.42
Girona, Y., Pratt, L. A., Kramarow, E. A., & Elgaddal, N. (2015). Hospitalization, readmission, and death experience of noninstitutionalized Medicare fee-for-service beneficiaries aged 65 and over. Centers for Disease Control and Prevention, 84, 1-23. Retrieved from https://www.cdc.gov/nchs/data/nhsr/nhsr084.pdf
Hansen, L. O., Young, R. S., Hinami, K., Leung, A., & Williams, M. V. (2011). Interventions to reduce 30-day rehospitalization: A systematic review. Annals of Internal Medicine, 155(8), 520-528. doi: 10.7326/0003-4819-155-8-201110180-00008.
Herrin, J., St. Andre, J., Kenward, K., Joshi, M. S., Audet, A-M., & Hines, S. C. (2014). Community Factors and Hospital Readmission Rates. Wiley Online Library. 50 (1), 20-39. doi:10.1111/1475-6773.12177
Joynt, K. E., & Jha, A. K. (2012). Thirty-Day Readmissions – Truth and Consequences. New England Journal of Medicine. 366, 1366-1369. doi:10.1056/NEJMp1201598
Joynt, K. E., & Jha, A. K. (2013a). A Path Forward on Medicare Readmissions. New England Journal of Medicine. 368, 1175-1177. doi:10.1056/NEJMp1300122
Joynt, K. E., & Jha, A. K. (2013b). Characteristics of Hospitals Receiving Penalties under the Hospital Readmissions Reduction Program. The Journal of the American Medical Association. 309(4), 342-343. doi:10.1001/jama.2012.94856
Kangovi, S., & Grande, D. (2011). Hospital readmissions – not just a measure of quality. The Journal of the American Medical Association, 306(15), 1688-1698. doi:10.1001/jama.20111515m
Kansagara, D., Englander, H., Salanitro, A., Kagen, D., Theobald, C., Freeman, M., & Kripalani, S. (2011). Risk prediction models for hospital readmission. A systematic review. Journal of the American Medical Association, 306(15):1688-1698. doi:10.1001/jama.2011.1515
Kripalani, S., Theobald, C. N., Anctil, B., & Vasilevskis, E. E. (2013). Reducing hospital readmission: Current strategies and future directions. The National Center for Biotechnology Information, 65, 471-485. doi:10.1146/annurev-med-022613-09041
Krumholz, H. M., Lin, Z., Keenan, P. S., Chen, J., Ross, J. S., Drye, E. E.,…Normand, S. L. (2013). The relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. The Journal of the American Medical Association, 306(16), 1796-1797. doi: 10.1001/jama.2011.1562
Merkow, R. P., Ju, M. H., Chung, J. W., Hall, B. L. Cohen, M. E., Williams, M. V.,…Bilimoria, K. Y. (2015). Underlying reasons associated with hospital readmission following surgery in the United States. The Journal of the American Medical Association, 306(16), 1796-1797. doi: 10.1001/jama.2011.1562
Miletic, K. G., Taylor, T. N., Martin, E. T., Vaidya, R., & Kaye, K. S. (2014). Readmissions after diagnosis of surgical site infection following knee and hip arthroplasty. Infection Control & Hospital Epidemiology, 35(02), 152-157.
Nedza, S., Fry, D. E., DesHarnais, S., Spencer, E., & Yep, P. (2016). Emergency department visits following joint replacement surgery in an era of mandatory bundled payments. Academic Emergency Medicine.
Rau, J. (2014). Medicare fines 2,610 hospitals in the third round of readmission penalties. Kaiser Health News Retrieved from https://khn.org/news/medicare-readmissions-penalties-2015/
Shah, T., Churpek, M. M., Coca, P. M., & Konetzka, R. T. (2014). Understanding why patients with COPD get readmitted. Journal Publications Chestnut, 147(5), 1219-1226. Retrieved from https://journal.publications.chestnet.org/
Sosunov, E. A., Egorova, N. N., Lin, H. M., McCardle, K., Sharma, V., Gelijns, A. C., & Moskowitz, A. J. (2016). The impact of hospital size on CMS hospital profiling. Medical care, 54(4), 373-379.
Swinburne, M., Garfield, K., & Wasserman, A. R. (2017). Reducing hospital readmissions: addressing the impact of food security and nutrition. The Journal of Law, Medicine & Ethics, 45(1_suppl), 86-89.
Tsai, T. C., Joynt, O., Gawande, A. A. & Jha, A. K. (2013). Variation in surgical readmission rates and quality of hospital care. New England Journal of Medicine, 369,1134-1142 doi:10.1056/NEJMsa1303118
Academic Master Education Team is a group of academic editors and subject specialists responsible for producing structured, research-backed essays across multiple disciplines. Each article is developed following Academic Master’s Editorial Policy and supported by credible academic references. The team ensures clarity, citation accuracy, and adherence to ethical academic writing standards
Content reviewed under Academic Master Editorial Policy.
