Eugene Litvak

Variability in Surgical Caseload and Access to Intensive Care Services

Background Variability in the demand for any service is a significant barrier to efficient distribution of limited resources. In health care, demand is often highly variable and access may be limited when peaks cannot be accommodated in a downsized care delivery system. Intensive care units may frequently present bottlenecks to patient flow, and saturation of these services limits a hospitals responsiveness to new emergencies. Methods Over a 1-yr period, information was collected prospectively on all requests for admission to the intensive care unit of a large, urban childrens hospital. Data included the nature of each request, as well as each patients final disposition. The daily variability of requests was then analyzed and related to the units ability to accommodate new admissions. Results Day-to-day demand for intensive care services was extremely variable. This variability was particularly high among patients undergoing scheduled surgical procedures, with variability of scheduled admissions exceeding that of emergencies. Peaks of demand were associated with diversion of patients both within the hospital (to off-service care sites) and to other institutions (ambulance diversions). Although emergency requests for admission outnumbered scheduled requests, diversion from the intensive care unit was better correlated with scheduled caseload (r = 0.542, P < 0.001) than with unscheduled volume (r = 0.255, P < 0.001). During the busiest periods, nearly 70% of all diversions were associated with variability in the scheduled caseload. Conclusions Variability in scheduled surgical caseload represents a potentially reducible source of stress on intensive care units in hospitals and throughout the healthcare delivery system generally. When uncontrolled, variability limits access to care and impairs overall responsiveness to emergencies.

Queuing Theory Accurately Models the Need for Critical Care Resources

Background: Allocation of scarce resources presents an increasing challenge to hospital administrators and health policy makers. Intensive care units can present bottlenecks within busy hospitals, but their expansion is costly and difficult to gauge. Although mathematical tools have been suggested for determining the proper number of intensive care beds necessary to serve a given demand, the performance of such models has not been prospectively evaluated over significant periods. Methods: The authors prospectively collected 2 years’ admission, discharge, and turn-away data in a busy, urban intensive care unit. Using queuing theory, they then constructed a mathematical model of patient flow, compared predictions from the model to observed performance of the unit, and explored the sensitivity of the model to changes in unit size. Results: The queuing model proved to be very accurate, with predicted admission turn-away rates correlating highly with those actually observed (correlation coefficient = 0.89). The model was useful in predicting both monthly responsiveness to changing demand (mean monthly difference between observed and predicted values, 0.4 ± 2.3%; range, 0–13%) and the overall 2-yr turn-away rate for the unit (21%vs. 22%). Both in practice and in simulation, turn-away rates increased exponentially when utilization exceeded 80–85%. Sensitivity analysis using the model revealed rapid and severe degradation of system performance with even the small changes in bed availability that might result from sudden staffing shortages or admission of patients with very long stays. Conclusions: The stochastic nature of patient flow may falsely lead health planners to underestimate resource needs in busy intensive care units. Although the nature of arrivals for intensive care deserves further study, when demand is random, queuing theory provides an accurate means of determining the appropriate supply of beds.

Managing Unnecessary Variability in Patient Demand to Reduce Nursing Stress and Improve Patient Safety

BACKGROUND Increases in adverse clinical outcomes have been documented when hospital nurse staffing is inadequate. Since most hospitals limit nurse staffing to levels for average rather than peak patient census, substantial census increases create serious potential stresses for both patients and nurses. By reducing unnecessary variability, hospitals can reduce many of these stresses and thereby improve patient safety and quality of care. THE SOURCE AND NATURE OF VARIABILITY IN DEMAND The variability in the daily patient census is a combination of the natural (uncontrollable) variability contributed by the emergency department and the artificial (potentially controllable) peaks and valleys of patient flow into the hospital fromelective admissions. Once artificial variability in demand is significantly reduced, a substantial portion of the peaks and valleys in census disappears; the remaining censsus variability is largely patient and disease driven. When artificial variability has been minimized, a hospital must have sufficient resources for the remaining patient-driven peaks in demand, over which it has no control, if it is to deliver an optimal level of care. DISCUSSION Study of operational issues in health care delivery, and acting on what is learned, is critical. Al forms of artificial variation in the demand and supply of health care services should be identified, and pilot programs to test operational changes should be conducted.

More Patients, Less Payment: Increasing Hospital Efficiency In The Aftermath Of Health Reform

A major issue for the US health care system will be accommodating the needs of the estimated thirty-two million Americans who will gain insurance coverage under the Affordable Care Act by 2019. For hospitals, a traditional response to this increased demand might be to add resources, such as more staff and beds. We argue that such actions would be unaffordable and unnecessary. Research has demonstrated that large gains in efficiency can be made through streamlining patient flow and redesigning care processes. We argue that once managed efficiently, US hospitals, on average, could achieve at least an 80-90 percent bed occupancy rate--at least 15 percent higher than the current level--without adding beds at capital costs of approximately

Screening for the Presence of a Disease by Pooling Sera Samples

1 million per bed. This article outlines a plan for hospitals to accommodate more patients without increasing beds or staff, and for policy makers to require hospitals to make these changes or provide incentives for them to do so.

Rethinking Rapid Response Teams

Abstract Screening of pooled urine samples was suggested during the Second World War as a method for reducing the cost of detecting syphilis in U.S. soldiers. Recently, pooling has been used in screening for human immunodeficiency virus (HIV) antibody to help curb the further spread of the virus. Pooling reduces the cost but also—and probably more importantly—offers a feasible way to lower the error rates associated with labeling samples when screening low-risk HIV populations. For example, given the limited precision of the presently available test kits, when the screened population has a prevalence of 4 per 1,000 (which is roughly the estimated U.S. prevalence), the probability that a sample labeled positive is antibody-free can be reduced from approximately 90% (if each sample is tested individually) to about 2%. Furthermore, screening pooled sera samples can also be used to reduce the probability that a sample labeled negative in fact has antibodies up to 40-fold in such a population—an important cons...

Whose Blood Is Safer?: The Effect of the Stage of the Epidemic on Screening for HIV

CURRENT DEBATE IN THE MEDICAL COMMUNITY centers on the benefits of rapid response teams (RRTs), hospital-based teams composed of clinicians with intensive care unit (ICU)–level clinical expertise. These teams rapidly respond when the condition of patients being cared for outside of the ICU suddenly deteriorates, and such patients often require transfer to ICUs. Those on one side of the debate suggest that RRTs save lives; this assertion is supported by common sense, numerous anecdotal reports, and some observational studies. Those on the other side of the debate suggest that preventing, recognizing, and treating deteriorating patients is common sense. How best to achieve this remains elusive based on systematic reviews, which have failed to show benefit of RRTs but note that RRT studies were often of poor quality and clinicians often failed to call an RRT when they should have, leading to uncertainty in the estimates of benefit. Proponents favor further research, encouraging hospitals to experiment with strategies such as RRTs, enhanced nurse staffing, or hospitalists who would respond to deteriorating patients, stressing prevention rather than recovery from deterioration. Those on both sides of the debate are united in their frustration that patients are needlessly experiencing morbidity and agree that preventing patients’ health from deteriorating is the optimal solution. The debate obscures a more fundamental question: why are RRTs needed in the first place? The answer seems to be simple. An RRT is needed when the condition of a patient who is receiving care in a medical/surgical unit deteriorates or requires ICU-level expertise to avoid further deterioration or even death. There are 2 reasons patients deteriorate. First, some deteriorate despite adequate clinical care. These patients would benefit from having an organized system to identify and treat patients whose conditions worsen, such as an RRT or code team. Second, patients deteriorate because of inadequate care; in other words, the level of care (eg, clinician training, staffing) provided to the patient in the inpatient unit is inadequate for the patient’s condition. Even though empirical evidence regarding the proportion of RRT calls caused by each of these reasons is lacking, the philosophy of RRTs is premised on the idea that current care is inadequate; therefore, introducing ICU-level care will benefit the patient. If current care is adequate, an RRT is not likely to make a difference. Underlying inadequate care is that patients have been admitted to a unit that provides inadequate care. A triage error or inability to admit or transfer a patient to the preferred unit is the main driver of patient misplacement. Underlying the triage error is the way patient flow is managed or mismanaged. Every physician and nurse would prefer that patients are cared for in a unit that can provide the appropriate level of care, where sufficient physician, nurse, and monitoring resources are available. Physicians commonly request that their patients remain in the ICU or are admitted to a specific nursing unit, often with monitored beds, believing care is better in some units than others. Intensive care units and monitored beds are scarce resources, demand for these resources periodically exceeds supply, and patients are often not admitted to these preferred units. This situation is especially problematic in hospitals without critical care physicians who use clearly defined protocols to coordinate the use of monitored beds. A common although often erroneous solution is to add more ICU and monitored beds. Even if the cost of adding a bed (about

Smoothing the Way to High Quality, Safety, and Economy

1 million capital for a regular inpatient bed) is ignored, experience suggests that adding more beds does not solve this problem. Eventually, demand for these beds will again exceed capacity. Why, then, is there a seemingly insufficient number of ICU and monitored beds? Why don’t hospitals define which patients should use ICU and monitored beds? One reason is that mismanaged patient flow in the form of artificial peaks does not allow compliance with any such definition. Despite average US hospital occupancy of 66% to 67%, hospitals are periodically overcrowded. The key word is “periodically.” Rarely are particular hospital units overcrowded 100% of the time, in which case more beds would be needed. Rather, these hospitals are typically overcrowded on cer-

Sleep Faster! (Somebody else needs your blanket...)

Background. With improvements in HIV antibody test (ELISA) performance, the win dow of time between infection and seroconversion becomes a major source of error in HIV screening. The authors examined its impact on the false-reassurance rate (FRR). Methods. Test sensitivity was modeled as the product of two factors: the in herent sensitivity (sensitivity when antibody is present) and the probability that antibody is present in infected blood. A model of HIV and AIDS incidence was used to derive an estimate of the probability of remaining in the seronegative window (pw ) among those who are infected. With plausible assumptions, this probability approaches 0.03. The FRR was then estimated as a function of the probability of remaining in the se ronegative window, the prevalence of HIV, and the inherent sensitivity of the ELISA test were estimated. Results. The FRRs for two blood donor groups, one with an HIV prevalence of 0.004 and a typical probability of remaining in the seronegative window (pw = 0.03) and the other with a higher prevalence of 0.017 but fewer donors in the window (pw = 0.003), are equal (140 per million donors) if the blood is negative on a single ELISA test. After two negative tests or a single test that can detect antibody more reliably, however, the FRR is much higher in the group with the higher pw (= 120 per million compared with 50 per million), because the greater numbers of donors in the window more than offsets the lower prevalence. Conclusions. With improvements in inherent sensitivity of ELISA by virtue of technical progress or retesting, the preva lence of HIV infection may no longer play the critical role in degrading the results of blood screening. As inherent test performance improves, tests are increasingly likely to miss infected blood because of the seronegative-window error rather than because of measurement error. Window error plays a proportionally greater role during the early stages of HIV dissemination in a population where the incidence of new HIV infection is high relative to the incidence of AIDS. These findings may explain, in part, the recent observation that cases of transfusion of contaminated blood often take place in areas where AIDS epidemics have started recently. They also suggest that the traditional strategy of soliciting blood donors from low-prevalence populations may not always be optimal, unless such populations are truly low-risk. Key words: HIV; AIDS; prevalence; incidence; sensitivity; ELISA; predicted values; protocols of screening; false-reassur ance rate. (Med Decis Making 1997;17:455-463)

Time series analysis of variables associated with daily mean emergency department length of stay

Artificial peaks and valleys in hospital demand, driven by planned surgeries, foster health care delivery that endangers patients, reduces access to care, puts pressure on clinicians at some times, and results in underutilization of health care resources at others.