Abstract
This study describes the degree to which blood draws occurred among hospitalized patients during traditional sleep hours and investigates trends over time.
Sleep is a key element of patient care and experience, and its deprivation among hospitalized patients is an important stressor that increases their risk of adverse events during and after hospitalization.1,2,3 Among the many sources of sleep interruption are early morning blood draws,4,5,6 which are often performed to inform decisions during morning rounds. We describe the degree to which blood draws occurred during traditional sleep hours and investigated trends over time.
Methods
We queried electronic health records for all blood draw events for adult patients hospitalized within both campuses (York Street and Saint Raphael) of the Yale New Haven Hospital (a large teaching hospital) between November 1, 2016, and October 31, 2019. We excluded blood draws performed in the emergency department, critical care units, and step-down units and those performed during the first 24 hours after admission to avoid nonroutine blood draws obtained during the initial evaluation. There were no local policies to specifically modify blood draw timing during the study period. Patients’ characteristics were obtained from the electronic health records. The institutional review board at Yale University exempted the study from review and need for patient consent.
We evaluated differences in timing of blood draws across patient subgroups using the Kruskal-Wallis test and compared the proportion of samples collected between 4:00 am and 6:59 am (early morning blood draws) by the characteristics of the patients using the χ2 test. In the Yale New Haven Hospital, primary clinical teams place blood draw orders specifying the exact desired timing of collection, which is set to 6:00 am as the default. We estimated the absolute difference in the percentage of early morning blood draws between November 2016 and October 2019 using a z score. All analyses were performed on sample-level data using Stata version 17.0 (StataCorp). A 2-sided P < .05 was considered statistically significant.
Results
There were 9 271 654 blood draws from 99 945 hospitalized patients during the study period; 3 598 500 samples were excluded. The final study sample included 5 676 092 blood draws from 79 347 patients.
Of the total samples, 2 206 410 (38.9%) were drawn between 4:00 am and 6:59 am (8.9% from 4:00 am to 4:59 am, 16.5% from 5:00 am to 5:59 am, and 13.5% from 6:00 am to 6:59 am) (Table). Outside the early morning time window, 20.7% of total blood draws were performed between 7:00 am and 11:59 am, 28.2% between 12:00 pm and 11:59 pm, and 12.2% between 12:00 am and 3:59 am. There were modest but statistically significant differences in the distribution of early morning blood draws across age groups, race and ethnicity, and sex (Table).
Table. Distribution of Timing of Blood Draw Overall and by Patient Characteristics (N = 5 676 092 Samples).
| Characteristics | Timing of early morning samples, median (IQR), ama | Early morning samples, %b | |||
|---|---|---|---|---|---|
| 4:00-4:59 am | 5:00-5:59 am | 6:00-6:59 am | 4:00-6:59 am | ||
| Total | 5:39 (5:03-6:13) | 8.9 | 16.5 | 13.5 | 38.9 |
| Age group, y | |||||
| 18 to 35 | 5:43 (5:07-6:13) | 8.6 | 17.7 | 15.0 | 41.3 |
| >35 to 65 | 5:38 (5:03-6:11) | 9.1 | 17.3 | 13.2 | 39.6 |
| >65 to 85 | 5:38 (5:02-6:13) | 8.9 | 15.7 | 13.0 | 37.6 |
| >85 | 5:45 (5:05-6:21) | 8.3 | 14.6 | 15.5 | 38.4 |
| Race and ethnicityc | |||||
| Non-Hispanic Asian | 5:38 (5:02-6:09) | 9.3 | 17.6 | 13.1 | 40.0 |
| Non-Hispanic Black | 5:42 (5:07-6:17) | 7.5 | 14.8 | 13.6 | 35.9 |
| Hispanic or Latino | 5:39 (5:04-6:12) | 8.3 | 16.3 | 12.8 | 37.4 |
| Non-Hispanic White | 5:38 (5:02-6:12) | 9.4 | 17.0 | 13.5 | 39.9 |
| Non-Hispanic other raced | 5:38 (5:04-6:10) | 8.9 | 17.3 | 13.3 | 39.5 |
| Sex | |||||
| Female | 5:39 (5:03-6:14) | 9.1 | 16.6 | 14.0 | 39.7 |
| Male | 5:38 (5:03-6:12) | 8.8 | 16.4 | 13.0 | 38.2 |
Defined as blood draws performed between 4:00 am and 6:59 am. The differences were tested using the Kruskal-Wallis test.
The differences were tested using the χ2 test. P < .001 for each category in each column.
Obtained from electronic health records and categorized into these 5 mutually exclusive groups.
Includes American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, other race or race not listed, unknown, and declined to answer.
From November 2016 through October 2019, the monthly proportion of blood samples drawn between 4:00 am and 6:59 am increased from 36.9% to 41.4% (change of 4.5 percentage points [95% CI, 4.1-4.9 percentage points]; P < .001]) (Figure). By hourly blocks, the proportion of samples drawn between 4:00 am and 4:59 am declined from 9.5% to 7.7% (change of −1.8 percentage points [95% CI, −2.1 to −1.6 percentage points]), the proportion drawn between 5:00 am and 5:59 am increased from 14.6% to 18.8% (change of 4.2 percentage points [95% CI, 3.9-4.5 percentage points]), and the proportion drawn between 6:00 am and 6:59 am increased from 12.8% to 14.9% (change of 2.1 percentage points [95% CI, 1.8-2.4 percentage points]) (P < .001 for all comparisons).
Figure. Temporal Trends in Early Morning Blood Draws Among Hospitalized Patients Between November 2016 and October 2019.
Early morning samples were defined as blood draws performed between 4:00 am and 6:59 am.
Discussion
In this study, the proportion of early morning blood draws was persistently high over 3 years, representing nearly 4 in 10 daily blood draws at the end of the study period. The timing distribution within early morning blood draws shifted to slightly later over the course of the study. This study, the largest on the timing of blood draws to our knowledge, suggests that early morning phlebotomies remain common among hospitalized patients despite growing evidence of the detrimental effects of sleep disruptions.
The modest timing differences across patient characteristics indicates the pervasive nature of early morning phlebotomy practices. Such practices are especially relevant for elderly patients, in whom sleep interruptions may increase the risk of delirium and posthospital syndrome.1
Early morning blood draws may be necessary to assess patients’ health status during morning rounds and occasionally to inform discharge decisions, but patient-centered care should consider strategies to limit nonurgent tests during sleep hours. Studies are needed to assess if improving sleep by limiting early morning blood draws and other interruptions could improve patient outcomes without untoward effects on quality of care.
Limitations include that the study used data from a single teaching hospital and that urgent early morning blood draws performed on medical and surgical floors could not be excluded from the analysis.
Section Editors: Jody W. Zylke, MD, Deputy Editor; Kristin Walter, MD, Senior Editor.
References
- 1.Krumholz HM. Post-hospital syndrome—an acquired, transient condition of generalized risk. N Engl J Med. 2013;368(2):100-102. doi: 10.1056/NEJMp1212324 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.DePietro RH, Knutson KL, Spampinato L, et al. Association between inpatient sleep loss and hyperglycemia of hospitalization. Diabetes Care. 2017;40(2):188-193. doi: 10.2337/dc16-1683 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Stewart NH, Arora VM. Sleep in hospitalized older adults. Sleep Med Clin. 2018;13(1):127-135. doi: 10.1016/j.jsmc.2017.09.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Grossman MN, Anderson SL, Worku A, et al. Awakenings? patient and hospital staff perceptions of nighttime disruptions and their effect on patient sleep. J Clin Sleep Med. 2017;13(2):301-306. doi: 10.5664/jcsm.6468 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wesselius HM, van den Ende ES, Alsma J, et al. ; Onderzoeks Consortium Acute Geneeskunde Acute Medicine Research Consortium . Quality and quantity of sleep and factors associated with sleep disturbance in hospitalized patients. JAMA Intern Med. 2018;178(9):1201-1208. doi: 10.1001/jamainternmed.2018.2669 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dobing S, Frolova N, McAlister F, Ringrose J. Sleep quality and factors influencing self-reported sleep duration and quality in the general internal medicine inpatient population. PLoS One. 2016;11(6):e0156735. doi: 10.1371/journal.pone.0156735 [DOI] [PMC free article] [PubMed] [Google Scholar]
