A properly-functioning human body relies on minerals capable of carrying an electric charge. Medical experts call these minerals `electrolytes’ and place potassium within this grouping.
Doctors often advise us to eat bananas and other foods rich in potassium, which helps stabilize the electrical activity in our hearts and assists in building muscle, among other functions. However people with certain medical conditions can end up with too much potassium in their bloodstreams, a condition known as hyperkalemia.
Weakness, unusual heart rhythms and nausea are all symptoms of the disease. Levels of potassium equal to, or greater than, 6.5mEq/L constitute potentially life-threatening hyperkalemia and require immediate hospitalization, according to the National Center for Biotechnology Information.
“For a healthy individual, it’s very difficult to become hyperkalemic,” says Oliver Lenz, M.D., a nephrologist with the University of Miami Health System. “It’s usually kidney disease that causes the trouble. “
Dr. Lenz says medical professionals treating hyperkalemic patients need to be “very thoroughly educated” about their patients’ diet. In Miami, “during the summer we see seasonal increases in hyperkalemia, because mangos are in season and they’re high in potassium,” Dr. Lenz says
Dexur’s researchers have taken a detailed look at hyperkalemia statistics coming out of Florida and the 49 other states, along with U.S. territories Puerto Rico, Guam, the U.S. Virgin Islands and the Northern Mariana Islands. The stats are associated with hospitalized hyperkalemia patients, and zero in on total inpatient discharges, total number of discharged hyperkalemia patients and the percentage of overall discharges that had hyperkalemia. Two other metrics highlighted are average length of stay difference with and without hyperkalemia at the diagnosis-related group (DRG) level, and average readmission difference with and without hyperkalemia at the DRG level.
“Newer agents such as patiromer sorbitex calcium and sodium zirconium cyclosilicate have been developed for long-term prevention and recurrence of hyperkalemia,” says Kristy Greene, PharmD, a Clinical Pharmacist Specialist with Emory University Hospital Midtown, in Atlanta. “Pharmacists can play an important role in identifying hospitalized patients early via clinical-decision tools, or by flagging their clinical records during medication profile review to prevent occurrences.”
Individuals being treated for heart failure are particularly susceptible to hyperkalemia, as are patients dealing with end-stage renal disease. Healthcare professions suggest that patients within these groups curtail dietary sources of potassium, and ask their physicians which medications tend to exacerbate hyperkalemia.
Sometimes, doctors will recommend a blood transfusion as a treatment option for someone with anemia. Typically, this is done in more extreme cases. Adding blood from a non-anemic person can help treat the anemia quickly and provide the body with a source of iron. However, the clinical impact on these patients is usually short-term and little data research has assessed the effect of blood transfusions on important health quality outcomes.
To assess the relationship between blood transfusions, anemia, and quality outcomes, Dexur data scientists analyzed data on the total number of patients with anemia and looked at what percentage of these patients had also received a blood transfusion. These scientists subsequently examined the difference in two quality measurement outcomes - average days in length of stay and average difference in readmission rates – between the group of anemic patients that received a blood transfusion and those that did not. All of this data was then organized by 54 separate US states and territories.
In general, the data showed that a sizable minority of patients with anemia received blood transfusions. For example. In California, 15.02% of the 2,932,623 inpatient anemia discharges also received a blood transfusion. In general, these patients stayed an average of 1.8 days longer than anemic patients who did not receive a blood transfusion. These patients also had higher rates of readmission: when Anemia and Blood Transfusions are both present, the readmission rate is 6.46% higher after accounting for case mix index, comorbidities, and complications at the DRG level.
Dr. Edmund K. Waller, MD, PhD, Professor in Hematology, Medical Oncology, Medicine, and Pathology at Emory University School of Medicine in Atlanta, GA, provided commentary on the data. He summarized the data as “showing that [approximately] 25%-30% of admissions include anemia as a diagnosis and about 50% of the patients with anemia receive a transfusion.” He noted that “Typically, patients who receive a transfusion have a more profound anemia than those who do not.”
In regard to the health outcomes, Dr. Waller said it is not surprising that patients with transfusions typically have longer LOS and more frequent readmission. “The transfusion takes about ½ a day, and patients with more profound anemia are usually sicker.”
Dr. Waller also commented on state-level variation. “I am not sure about why there is variation by state. It could be that some states admit less sick patients so that fraction of patients with anemia is less.” He went on to elaborate “however, it appears that there is no strong correlation between the fraction of admissions with anemia, and the fraction of admissions with anemia and transfusion, suggesting that different states may have different transfusion practices.” He also noted that there is no relationship between the fraction of anemic patients who receive transfusions and the total number of admissions by state.
By David Reilly
Methicillin-resistant staphylococcus aureas (MRSA) can lead to endocarditis, osteomyelitis, and septicemia in severe cases.1 And for as many as 11,000 people in the United States each year, MRSA can also be fatal.2, 3
For many years, the number of patients with MRSA infection in US hospitals continued to rise dramatically—from less than 2,000 cases in 1993 to approximately 368,600 cases in 2005.4 But as awareness has grown and protocols for reducing the transmission of MRSA have been implemented, significant inroads have already been made for its reduction. According to Centers for Disease Control data, the incidence of hospital-acquired MRSA fell by as much as 54% from 2005 to 2011 while the number of annual MRSA-associated deaths reduced by 9,000 over the same period.2, 3
In addition to reducing considerable morbidity and mortality, efforts to curb MRSA can also bring about significant reductions in healthcare costs. MRSA infection is strongly associated with increased length of stay (LOS) in hospital.4-6 For patients without MRSA infection, the average LOS is 4.6 days incurring an average of $7,600 in total hospital costs.4 For patients with MRSA infection, LOS more than doubles to 10.0 days at an average cost of $14,000.4
MRSA infection is also associated with increased risk of hospital readmission.7 According to a large study tracking data for 136,513 patients across 8 years, patients with a MRSA-positive culture more than 48 hours after hospital admittance have a 40% greater risk of being readmitted to hospital within a year compared to patients with a MRSA-negative culture (HR, 1.40; 95% CI, 1.33-1.46).7
Across the nation, institutions have taken various approaches against MRSA, with differing rates of success. According to a recent analysis from Dexur which compiled averages for excess LOS due to MRSA and MRSA-associated readmissions for each state, Rhode Island has one of the lowest differences in average LOS of just 1.63 days and also has one of the lowest differences in average excess readmission. So how does Rhode Island fare so well? According to the Rhode Island Department of Health (DOH), the state is vigilant in performing inspections of residential healthcare facilities and hospitals, tracking infection outbreaks, and monitoring rates of infection in hospitals as part of a broader Healthcare Quality measurement initiative.8
According to the Dexur analysis, New Hampshire also had one of the lowest LOS differences, with an average of just 1.95 excess days. A closer look at the Dexur data identifies St. Joseph Hospital of Nashua as having the lowest average MRSA-associated LOS in the entire state, with a difference of just 0.34 days compared to patients without MRSA. So Dexur reached out to their infection preventionist, Ashley Conley, MS, CPH, CHEP for her insights.
One of the most successful initiatives for primary prevention at St. Joseph Hospital involved reaching out to the community—collaborating with various groups to educate athletes, schools, and other at-risk sub-populations on MRSA, its symptoms, and its prevention because as Ms. Conley put it, “infectious diseases don’t stop at the hospital walls.” This initiative alone appears to have been a great success, with the number of MRSA cases reducing dramatically in the five years since its introduction.
A robust antibiotic stewardship program serves as the foundation for addressing existing infection. Ms. Conley explained the key tenets of the program as, “ensuring we obtain the appropriate cultures and prescribe the right antibiotic at the right time.” This is complemented by an attention to recognizing risk factors, testing for colonization in at-risk patients, and using appropriate isolation. The hospital also seizes the opportunity to further heighten awareness by educating not only MRSA patients but also the family members who visit them. Through their dedication to educational initiatives, it seems St. Joseph Hospital is empowering an entire community to control infection. It’s an approach delivering impressive results and one that other institutions could learn from.
1. Sutton J, Steiner C. Hospital-, Health Care-, and Community-Acquired MRSA: Estimates From California Hospitals, 2013: Statistical Brief# 212. 2006.
2. Centres for Disease Control and Prevention. Antibiotic resistance threats in the United States, 2013: Centres for Disease Control and Prevention, US Department of Health and Human Services; 2013.
3. Dantes R, Mu Y, Belflower R, Aragon D, Dumyati G, Harrison LH, et al. National burden of invasive methicillin-resistant Staphylococcus aureus infections, United States, 2011. JAMA internal medicine. 2013;173(21):1970-8.
4. Elixhauser A, Steiner C. Infections with methicillin-resistant Staphylococcus aureus (MRSA) in US hospitals, 1993–2005. 2007.
5. Gilligan P, Quirke M, Winder S, Humphreys H. Impact of admission screening for meticillin-resistant Staphylococcus aureus on the length of stay in an emergency department. Journal of Hospital Infection. 2010;75(2):99-102.
6. Macedo-Vinas M, De Angelis G, Fankhauser C, Safran E, Schrenzel J, Pittet D, et al., editors. Excess length of stay due to methicillin-resistant Staphylococcus aureus (MRSA) infection at a large Swiss hospital estimated by multi-state modelling. BMC Proceedings; 2011: Springer.
7. Emerson CB, Eyzaguirre LM, Albrecht JS, Comer AC, Harris AD, Furuno JP. Healthcare-associated infection and hospital readmission. Infection Control & Hospital Epidemiology. 2012;33(6):539-44.
8. State of Rhode Island Department of Health (DOH). Healthcare-Acquired Infections: State of Rhode Island DOH; 2017 [Available from: http://www.health.ri.gov/diseases/healthcareacquiredinfections/.
Nearly a decade ago, the surgeon general published a call to action to improve prevention and treatment of deep vein thrombosis (DVT) a condition in which a blood clot forms in a patient’s legs and has the potential to break loose, travel toward the lungs, and lead to a life-threatening pulmonary embolism (PE.) “Why do DVT and PE remain such serious problems, particularly given the availability of effective strategies for preventing and minimizing them? The answer lies primarily in the failure to consistently use evidence-based interventions in those high-risk individuals who need them.” The summary stated.
In 2017, DVT remains remains a massive problem in healthcare – one that, by some estimates costs the country tens of billions of dollars and more importantly, tens of thousands of lives. The majority of patients who experience DVT suffer from it after a surgery. Two of the major risk factors for DVT are undergoing surgery and having previously experienced DVT. A recent analysis from Dexur showed that throughout the nation, a patient who suffers from DVT spends two and a half more days in the hospital than one who didn’t, and is readmitted to the hospital 2.18 percent more often. The data however, varied widely from hospital to hospital, potentially reflecting varied approaches to balancing effective prevention and treatment while minimizing cost.
One strategy to decrease the toll of excess hospitalization for those hospitals that can adopt it, is treating DVT patients in an outpatient setting. Alex Spyropoulos, MD, Professor at Hofstra Northwell School of Medicine, helped set up one of the first outpatient clinics for treating these patients in the 1990s. Dr. Spyropoulos explained to Dexur that “We were doing stuff that’s still not being done now, 20 years later. But, to set up an anticoagulation clinic takes a lot of money, and you don't get that much in terms of revenues. Hospitals have to buy in that this is a huge quality issue.” He has worked on several studies estimating the costs of DVT, one of which included developing a model of the overall economic burden including treatment and loss of work for patients, which predicted DVT costs the United States tens of billions of dollars.
The trend of outpatient anticoagulation treatment is catching on in some hospitals. “It became apparent that we were keeping patients in the hospital for the sole purpose of waiting for their blood thinner therapy to become therapeutic.” Said Adam Porath, PharmD, BCACP, BCPS-AQ Cardiology Ambulatory Pharmacy Manager at Renown Health. Six years ago, he developed the outpatient service to manage DVT patients. Last month, at the American society of Health System Pharmacist meeting, the group presented a study that showed they were able to decrease both length of stay and the overall cost of care for their DVT patients using the outpatient protocol.
But financial investment isn’t the only roadblock to setting up an effective anticoagulation clinic. “It’s got to be a local solution” said Spyropoulos. Much of the feedback Porath received after presenting Renown’s results was that some hospitals aren’t designed to provide this type of integrated care. The most important part is ensuring that patients get the treatment that they need for the appropriate length of time – whether it’s in the hospital or not. The cost of treatment along with the cost to stay in the hospital to receive that treatment is overwhelming, so the patient’s hospital stays are getting shorter. But that’s where the breakdown can happen. The patients aren’t transferred home or to long term care with prophylaxis, and end up sicker. “All of sudden, patients are not getting the appropriate duration of thromboprophylaxis.” he explained, “Unfortunately, we're actually going backwards.”
Researchers in the past have studied associations of Anemia with health outcomes such as LOS & Readmission rates (see 4 studies listed at the end of the article). Similarly, we examined our database to analyze the impact on Length of Stay (LOS) when Acute Posthemorrhagic Anemia was present for major Ortho & Colorectal procedures.
The results are shown in the below chart & table.
|Category||DRG||Avg. LOS with Acute Posthemorrhagic Anemia||Avg. LOS without Acute Posthemorrhagic Anemia||Avg. LOS Difference|
|Ortho||470 - MAJOR JOINT REPLACEMENT OR REATTACHMENT OF LOWER EXTREMITY W/O MCC||3||2||1|
|Ortho||483 - MAJOR JOINT/LIMB REATTACHMENT PROCEDURE OF UPPER EXTREMITIES||3||1||2|
|Ortho||469 - MAJOR JOINT REPLACEMENT OR REATTACHMENT OF LOWER EXTREMITY W MCC||7||6||1|
|Colorectal||330 - MAJOR SMALL & LARGE BOWEL PROCEDURES W CC||8||7||1|
|Colorectal||329 - MAJOR SMALL & LARGE BOWEL PROCEDURES W MCC||16||12||4|
As the above data table shows, Ortho & Colorectal DRG procedure groups have a 1-4 day difference when Acute Posthemorrhagic Anemia was present versus when it was not present.
Previous papers studying the impact of Anemia on Quality & Economic metrics:
One of the most common topics when discussing length of stay (LOS) metrics is the myriad of ways one can reduce LOS while other metrics such as Skilled Nursing Facility (SNF) discharge rates & readmittance rates are adversely impacted. Specifically, regarding SNF Discharge rates, there is a perception that there is a correlation between lower LOS metrics & higher SNF discharge rates.
We decided to study if this perception was reflected in the data. We analyzed all discharges between July 2015 & June 2016 in the Medicare claims database & filtered only those discharges for DRG 470 - Joint Replacement. For background, DRG 470 is one of the main procedure groups for Comprehensive Care for Joint Replacement (CJR) that is part of the Bundled Payments initiative OF CMS. SNF discharge rates are closely tracked within the CJR & BPCI framework to ensure that quality outcomes are not being adversely impacted while economic metrics improve. We then filtered under the criteria of a minimum of 100 Medicare discharges in the 12 month period between July 2015 to June 2016 and were left with data on 1,506 hospitals.
While the work performed to gather the data for analysis was complex, the end result can be shown through a simple scatter plot (shown below) for the 1,506 Hospitals.
America’s problem with the Opioid Crisis is well known & there is hardly a day that goes by without research & news on how this crisis can be combatted. According to the CDC, more than half a million people have died between 2000 to 2015 & about 91 people die everyday. We analyzed claims within the Medicare database to understand key opioid abuse hospitalizations & readmittance rates within a 30-60-90 day window.
The below chart for the two ICD codes related to opioid poisoning shows readmittance rates of about 27-29% within 30 days and 42% within 90 days.
The above data is from discharges between Jan to June 2015 from CMS Medicare Inpatient claims data. These statistics can be further augmented by data from from the HCUP 2014 database for the two ICD codes (which is shown in the below table).
|ICD Code||ICD Code Description||Length of Stay||Average Charges||Average Costs|
As is evident from the above table, each stay costs about $10,000 per stay & every extra day in the hospital costs the system about $2500/day. These costs are further compounded by more than 40% of these patients returning to the hospitals for treatment. To improve care for patients with potential Opioid Addiction, treatment should be considered longitudinally across a longer time frame so that both clinical & economic outcomes are positively improved.
Dexur utilizes advanced statistical techniques & large scale data sets such as medical claims & other data sources to show in depth insights at the hospital, disease & practice level. To get more detailed insights, please contact us here.
Length of Stay (LOS) is perhaps one of the most closely watched metrics in inpatient Hospital settings. From Wikipedia on what LOS means:
A common statistic associated with length of stay is the average length of stay (ALOS), a mean calculated by dividing the sum of inpatient days by the number of patients admissions with the same diagnosis-related group classification. Length of stay (LOS) is a term to describe the duration of a single episode of hospitalization. Inpatient days are calculated by subtracting day of admission from day of discharge.
LOS is a big focus area for Insurance companies & hospitals. For example, Medicare, through its Bundled Payments for Care Improvement (BPCI) Initiative, aims to pay a flat fee for a single type of surgery such as Knee Replacement. In that scenario, Hospitals are extremely motivated to reduce the LOS for a single surgery since that reduces the costs of the Hospital while keeping the same fee payment. Given that context, predictive capabilities around LOS are an extremely important area for innovation.
Dexur analyzes large scale medical claims data sets to identify LOS by discharge & diagnoses area for all hospitals. We have created a large and enriched LOS data set based on claims for all hospitals to aid in the development of machine learning models. If you are a healthcare researcher & want access to these data sets, please contact us & we can collaborate on a project. A simple chart based on the data set showing the top Discharge Groups by LoS at Mayo Clinic at Rochester is given below & the details can be seen here.
In addition, we have also shared 5 Machine Learning studies that try to predict & understand LOS in Hospitals:
IMPROVED PREDICTION OF HOSPITAL LENGTH OF STAY FOR SEVERE INJURY: There are limited beds in hospital trauma wards, and yet there is a constant demand for these beds by the inflow of severely injured patients. Many patients are initially allocated to these beds when they could be better treated in another specialised ward. If we could accurately classify patients with hospital length of stay (LOS) of 2 days or less versus those who require longer stays, we could make a more informed decision whether or not to place them in another ward when they are admitted, rather than wasting time and resources transferring them to another ward later. The study was conducted on two datasets: one with 2546 records from the Trauma Services Centre at the Royal Prince Alfred Hospital in Sydney, consisting of trauma patients admitted to the centre between 2007–11; the other from the Hospital das Foras Armadas in Portugal with 17546 records collected from 2000–13 and covering a wide range of medical diagnoses. The authors investigate feature transformation and selection techniques in the construction of a LOS prediction model for trauma patients. They also apply and evaluate a comprehensive range of classification algorithms on data from the trauma domain as well as from a general hospital setting. In addition, the authors propose a new nearest neighbour (NN) algorithm, ranked NN, which takes into account the predictive relevance of features when computing the distance to the nearest neighbors.
Length of Stay Prediction and Analysis through a Growing Neural Gas Model: In this work a novel unsupervised LoS prediction model is presented which performs better than other ones commonly used in this kind of problem. The developed model detects autonomously the subset of non-class attributes to be considered in these classification tasks, and the structure of the trained self organizing network can be analysed in order to extract the main factors leading to the overcoming of regional LoS threshold. The Growing Neural Gas (GNG) model is capable in identifying exactly the local dimension of the input space. The paper explains how the authors obtained a higher accurate prediction by the use of GNG in comparison with other algorithms which are commonly used in these kind of problems.
MACHINE LEARNING TECHNIQUES FOR PREDICTING HOSPITAL LENGTH OF STAY IN PENNSYLVANIA FEDERAL AND SPECIALTY HOSPITALS: For inpatient care units, two variables play an important role in determining hospital resource utilization. The first variable is predicting a patient’s hospital length of stay (LOS), and second variable is predicting readmissions [Kelly et al. (2013)]. Ideally, a hospital must minimize both variables to provide high-quality healthcare and improve resource utilization. Predicting hospital LOS allows a hospital to predict discharge dates for a patient admitted to the hospital, which in turn allows improved scheduling of elective admissions leading to reduce variance in hospital bed occupancies. Predicting LOS also allows a hospital to scale its capacity during its longterm strategic planning. In this paper, we compare three different machine learning techniques for predicting length of stay (LOS) in Pennsylvania Federal and Specialty hospitals. Using the real-world data on 88 hospitals, the authors compare the performances of three different machine learning techniques—Classification and Regression Tree (CART), Chi-Square Automatic Interaction Detection (CHAID) and Support Vector Regression (SVR)—and find that there is no significant difference in performances of these three techniques. However, CART provides a decision tree that is easy to understand and interpret. The results from CART indicate that psychiatric care hospitals typically have higher LOS than nonpsychiatric care hospitals. For non-psychiatric care hospitals, the LOS depends on hospital capacity (beds staffed) with larger hospitals with beds staffed over 329 having average LOS of 13 weeks vs. smaller hospitals with average LOS of about 3 weeks.
A Comparison of Supervised Machine Learning Techniques for Predicting Short-Term In-Hospital Length of Stay Among Diabetic Patients: Due to the growing number of hospitalized diabetic patients, predicting the average length of stay (LOS) has become increasingly important for both resource planning and effective admission scheduling. Obtaining LOS estimates is useful for planning future bed usage, determining specialists for patients with multiple diagnoses, determining health insurance schemes and reimbursement systems in the private sector, planning discharge dates for elderly patients, and allowing families to better plan for the return of their relatives. In this paper, the authors compare and discuss the performance of various supervised machine learning algorithms (i.e., multiple linear regression, support vector machines, multi-task learning, and random forests) for predicting long versus short-term length of stay of hospitalized diabetic patients.
Real-time prediction of inpatient length of stay for discharge prioritization: Hospitals are challenged to provide timely patient care while maintaining high resource utilization. This has prompted hospital initiatives to increase patient flow and minimize nonvalue added care time. Real-time demand capacity management (RTDC) is one such initiative whereby clinicians convene each morning to predict patients able to leave the same day and prioritize their remaining tasks for early discharge. Our objective is to automate and improve these discharge predictions by applying supervised machine learning methods to readily available health information. The authors use supervised machine learning methods to predict patients’ likelihood of discharge by 2 p.m. and by midnight each day for an inpatient medical unit. Using data collected over 8000 patient stays and 20 000 patient days, the predictive performance of the model is compared to clinicians using sensitivity, specificity, Youden’s Index (i.e., sensitivity þ specificity – 1), and aggregate accuracy measures. The model compared to clinician predictions demonstrated significantly higher sensitivity (P<.01), lower specificity (P<.01), and a comparable Youden Index (P>.10). Early discharges were less predictable than midnight discharges. The model was more accurate than clinicians in predicting the total number of daily discharges and capable of ranking patients closest to future discharge. Conclusions There is potential to use readily available health information to predict daily patient discharges with accuracies comparable to clinician predictions. This approach may be used to automate and support daily RTDC predictions aimed at improving patient flow.