Background: Sleep is generally poor in hospitalized patients and sleep deprivation has many harmful effects, including delirium, insulin resistance, hypertension, and impaired function of the immune system. In this quality improvement (QI) project, our goal was to delay early morning phlebotomy blood collections to improve sleep for our patients, while not negatively impacting hospitalists’ perception of care delivery nor increasing workload for physicians and nurses.
Methods: Nursing workload was determined by comparing the number of morning blood collections performed by nursing instead of phlebotomy. Physician workload was assessed by the number of early morning pages received by night hospitalists, as anecdotally hospitalists reported receiving numerous pages for routine electrolyte supplementation prior to the start time for the day hospitalists.
At a large academic medical center this QI project was undertaken on a single medical floor to start morning phlebotomy collections no earlier than 5am. Approximately three months of lab and pager data pre- and post- intervention were collected. A physician survey was conducted post-intervention to assess the perceived effect of this change among the hospitalist group. A patient survey was also collected pre- and post-intervention to compare sleep at home versus in the hospital. This QI project was deemed exempt upon review by the Institutional Review Board at the University of Kentucky.
Results: Phlebotomy collection time was delayed by 155 min to 7:17 a.m. (p = <0.001) with no change in percentage of nurse collected labs (p = 0.749) nor pages received to the night hospitalist (p = 0.788). The hospitalist survey had a response rate of 52% (23/44) with 83% (19/23) reporting that the pilot did not require the daytime hospitalist to change their usual rounding routine. 74% (17/23) of hospitalists also reported believing it was beneficial for their patients to delay the morning lab phlebotomy, and 65% (15/23) ultimately felt the project should be continued. In the patient survey, there was no significant difference in sleep or interruptions in sleep in the hospital pre or post-intervention.
Conclusions: In this QI project, phlebotomy collection time was delayed by more than two-and-a-half hours. Despite this later collection time, we found no associated increase in nursing work, contrary to initial concern that physicians would request the nurses to perform more of the phlebotomies to obtain lab results more quickly. Initial hopes for decreased pages to the night hospitalists were not realized, though the baseline number was less than expected. The majority of physicians who completed the survey did not perceive that the delayed lab collection negatively affected their work and supported continuation of the pilot. However, we found no statistical change in patient’s perception of sleep in the hospital after this intervention.