Intelligence Unit Special Reports Special Events Subscribe Sponsored Departments Follow Us

Twitter Facebook LinkedIn RSS

Patient Safety: Does Variability in Admissions Matter?

Brad Prenney, MS, MPA, and Kathleen Kerwin Fuda, PhD, for HealthLeaders Media, December 20, 2007
Patient safety is a major concern to hospital leaders, clinicians, policymakers and patients alike. Projects to increase patient safety and the overall quality of healthcare typically focus on improvements to specific clinical processes. Although these efforts are important and valuable, alone they are not sufficient to ensure patient safety. The variability in the number of patient admissions from day to day, and the resulting swings in census and provider workload, need to be addressed.

Fortunately, there are solutions available to hospital managers to better manage these swings and to improve patient flow. Use of Variability Methodology will minimize unnecessary fluctuations in admissions and census, and improve management of those that are unavoidable.

The key principles of Variability Methodology are:

  • identification of the sources and types of variability. With regard to the number of patient admissions, there are two types of variability: artificial and natural.
    1. Artificial variability is variability that is created by the hospital itself, and is not due to patients' clinical needs. Poor scheduling of elective admissions is typically the main source of artificial variability in patient flow, which is a much larger problem than is generally recognized.
    2. Natural variability refers to variability that is inherent in the incidence of illnesses or injuries requiring patient admissions.
    1. elimination of artificial variability in admissions. Frequently this is done through improved surgical scheduling, although other sources of patient admissions may need to be addressed as well, e.g., obstetrical admissions.
    2. better management of the natural variability in emergency admissions. Through tools from operations research, such as queuing models, fixed capacity may be efficiently matched to variable demand.
    3. separation of patient subpopulations, as volume levels permit, to increase efficiency and improve quality. For example, separate OR resources generally should be dedicated to handle emergent and urgent cases, as opposed to elective case loads.

    Safety impacts of variability
    The impacts of variability in admissions and census on patient safety and quality of care are wide-ranging. Effects can include:

    • over-stressing of the work force, resulting in errors and mistakes.
    • substitution of less appropriate care for that which would normally be offered, e.g., patients placed "off-service";
    • outright rejection of patients seeking care, e.g., ambulance diversions or ED patients leaving without being seen;
    • waits and delays in receiving care;
    • cancellation of procedures.

    Nurse staffing is typically determined by average census level. However, during peaks in census, nurses are forced to care for more patients than planned. The adverse impact of increasing patient to nurse ratios on both risk of patient mortality and quality of care received has been well documented. Unfortunately, our work with hospitals has shown that these ratios regularly vary by as much as 100 percent from day to day due to changing patient census, resulting in excessive stress on staff and setting up an environment ripe for medical errors and quality problems.

    An important quality problem that deserves more attention is misplacement of patients once admitted to the hospital, due to shortage of beds in the preferred unit. Compromises are made, and patients are sent to a bed "off-service." This practice always denies patients the benefits of care in a unit specializing in their particular health problems, and it is particularly dangerous when patients do not receive the intensity of care that is needed, for example, when a bed in the ICU is unavailable for a patient needing one.

    Problems caused by variable patient demand affect the emergency department most visibly. The inability to move admitted ED patients to floor beds is the principal cause of ED overcrowding. This is most often a result of ED-admitted patients competing with elective surgical patients for the same beds. To cope with overcrowding, the hospital may divert ambulances to other hospitals, thus rejecting patients seeking care there. Others are boarded in the ED for hours (and sometimes days) waiting to be transferred to an inpatient bed.

    No service is immune to the impact of variable patient demand. In the operating room, urgent and emergent surgeries are delayed or end up bumping elective surgeries as cases compete for a room and staffing. Unpredictable peaks in demand result in the OR and PACU working overtime and under stress to get cases done, situations that not only increase the risk of errors but that affect patient and caregiver satisfaction. Even ancillary departments such as radiology or therapy services experience increased waiting times for services when patient demand peaks.

    What Hospital Leaders Can Do
    Hospitals that have applied Variability Methodology to their operations to improve patient flow have experienced major positive improvements to both their operations and quality of patient care.

    • At Cincinnati Children's Hospital Medical Center, separation of OR resources for emergent/urgent vs. elective surgical cases, a fundamental principle of Variability Methodology, resulted in a 28 percent reduction in waiting time for performing emergent/urgent cases even though volume increased 24 percent. In addition, OR overtime was reduced by 57 percent. A survey of CCHMC clinicians after implementation indicated wide scale satisfaction with the changes, especially among surgeons.
    • At Boston Medical Center, similar separation of resources for emergent/urgent cases reduced "bumping" of elective surgical cases, formerly required to accommodate them, by over 99 percent. Variable patient flow to the step-down unit was reduced by half, improving staff satisfaction. In addition, average Nursing Hours per Patient Day (NHPPD) were reduced by .6 hour, by reduction of periods of low census in the unit, saving $130,000 per year. The ED benefited as well as ambulance diversions declined, wait times in the ED were reduced by 1/3, and throughput time decreased by 45 minutes, allowing more patients to be seen.

    Improvement of hospital-wide patient flow requires commitment and leadership from senior hospital leaders and involvement of clinicians and managers from the OR (particularly surgeons), the ED, and inpatient units. Reducing the variability of elective admissions, typically through the OR, and right-sizing of capacity to best manage the remaining variability due to emergent admissions, often requires significant changes. However, proper management of variability in patient demand is essential to improving patient safety and quality of care, while promising significant gains in hospital efficiency as well.

    For more detailed information about Variability Methodology, visit www.bu.edu/mvp.


    Kathleen Kerwin Fuda, Ph.D, is a consultant for PatientFlow Technology, Inc., and data analysis manager of the Managing of Variability Program at the Boston University Health Policy Institute. Brad Prenney, MS, MPA, is a consultant for PatientFlow Technology and Deputy Director of the Management of Variability in Health Care Delivery at the Boston University Health Policy Institute.
    For information on how you can contribute to HealthLeaders Media online, please read our Editorial Guidelines.

      Comments are moderated. Please be patient.