Frequent measurement of vital signs has been associated with disruptions to sleep and increased nursing workload. Since vital signs are often measured at the same frequency regardless of patient acuity, there may be inappropriate prioritization of limited resources. We sought to understand what hospitalists report as the default frequency of routine vital sign measurement in hospitalized pediatric patients at academic institutions.
We surveyed pediatric hospital medicine leadership at Association of American Medical Colleges-affiliated medical schools on their perception of routine vital signs in general medicine inpatients.
Survey requests were sent to individuals representing 140 unique hospitals. Responses were received from 74 hospitalists, representing a 53% response rate. Routine vitals were most commonly characterized as those collected every 4 hours (78%; 95% confidence interval, 67%–87%), though at least 1 in 5 hospitalists reported obtaining all or select vital signs (eg, blood pressure) less frequently. Strategies to decrease vital sign frequency varied.
Our results suggest routine vital signs are not a normative concept across all patient populations in pediatrics. We further identify several conditions under which deviation from routines are sanctioned.
An important purpose of hospitalization is to monitor patients for the risk of clinical deterioration. A mainstay of this monitoring is through the recording of serial vital signs, generally considered to include heart rate, respiratory rate, blood pressure, and temperature. Oximetry is variably included.1 Although scheduled routine monitoring is intended for earlier detection of patient deterioration, and often considered to be normative, frequent measurement of all vital signs is not without consequence. Vital signs are consistently identified as a major disruptor to patient sleep in the hospitalized setting.2–4 Blood pressure, which is arguably one of the more disruptive vital signs, is frequently measured inaccurately, which can lead to over and underdiagnosis of hypertension.5,6 Vital sign measurement is additionally a considerable source of nursing workload,1,7 regardless of patient acuity.8 It is also known that vital signs are, in practice, often measured at a different frequency than the one intended by the ordering provider.9
Quality improvement projects aiming to improve sleep, address nurse overwork, facilitate the appropriate prioritization of nursing tasks, and/or improve the value of care provided to patients may attempt to alter the default practice of obtaining frequent vital signs in all patients. To successfully address the problem, leaders must first understand the frequency of patient monitoring providers consider normative. We surveyed hospitalists to characterize what they report to be routine vital sign frequency at their institutions, as well as conditions under which deviation from routine was sanctioned.
Methods
We designed a survey instrument following the methodology described in Ziniel et al.10 Investigators with training in survey design were consulted to establish content validity. Cognitive pretesting was completed with additional individuals who represented the survey’s target population.
Our sampling frame included 140 pediatric hospital medicine program leaders at programs affiliated with the Association of American Medical Colleges (AAMC) list of Liaison Committee on Medical Education-accredited United States medical schools, using methods previously described by Allan et al.11 Medical schools in Puerto Rico and those with pending or provisional accreditation at the time the list was created (2020) were not included. Two institutions had >1 hospital medicine division within the same hospital. In these instances, only 1 division leader was surveyed.
Our web-based survey link was e-mailed to participants in August and September 2021. Hospitalists were asked about the default frequency of routine vital signs on general medical patients at the primary clinical teaching site at their institutions. To assess response rate, respondents specified the names of their hospitals. A maximum of 1 reminder e-mail was sent. Participants had the option to provide free-text comments or clarifications at the end of the survey.
The total number of hospital beds at each site was collected from the Children’s Hospital Association (CHA) database. If the hospital was not part of this database, this information was collected from the hospital’s web site (n = 13). We were unable to determine bed size in 3 hospitals in the respondent and 3 hospitals in the nonrespondent group. These hospitals were excluded from this calculation. The designation of pediatric level I trauma center was collected from the American College of Surgeons Committee on Trauma database. These metrics were used as broad proxies for patient volume and acuity of a hospital.
Descriptive statistics using 95% confidence intervals (CIs) for proportions are presented. Differences between continuous variables were evaluated using t tests. Differences between categorical variables were evaluated using a 2-sample proportion z-test.
This study was approved as nonhuman subjects research through the university’s institutional review board.
Results
Survey requests were sent to individuals representing 140 unique hospitals with affiliations with an AAMC-accredited medical school with inpatient pediatric programs. Responses were received from 74 hospitalists, a 53% response rate. One respondent answered “don’t know” to routine vital sign and routine pulse oximetry frequency and was thus excluded from associated calculations. There were no differences in bed size, trauma center designation, CHA membership, or geographic region found between respondents and nonrespondents (Table 1).
Demographic Characteristics of Survey Respondents Versus Nonrespondents
| . | Respondents (n = 74) . | Nonrespondents (n = 66) . | Difference (95% CI) . |
|---|---|---|---|
| Average total beds | 208 | 176 | 32 (−14 to 77) |
| Level I trauma center | 54% (n = 40) | 41% (n = 27) | 13% (−11% to 37%) |
| CHA member | 89% (n = 66) | 83% (n = 55) | 6% (−1% to 18%) |
| Region | |||
| Northeast | 24% (n = 18) | 24% (n = 16) | 0 (−28% to 28%) |
| South | 37% (n = 27) | 42% (n = 28) | −5% (−21% to 31%) |
| Midwest | 22% (n = 16) | 26% (n = 17) | −4% (−25% to 33%) |
| West | 18% (n = 13) | 8% (n = 5) | 10% (−27% to 47%) |
| . | Respondents (n = 74) . | Nonrespondents (n = 66) . | Difference (95% CI) . |
|---|---|---|---|
| Average total beds | 208 | 176 | 32 (−14 to 77) |
| Level I trauma center | 54% (n = 40) | 41% (n = 27) | 13% (−11% to 37%) |
| CHA member | 89% (n = 66) | 83% (n = 55) | 6% (−1% to 18%) |
| Region | |||
| Northeast | 24% (n = 18) | 24% (n = 16) | 0 (−28% to 28%) |
| South | 37% (n = 27) | 42% (n = 28) | −5% (−21% to 31%) |
| Midwest | 22% (n = 16) | 26% (n = 17) | −4% (−25% to 33%) |
| West | 18% (n = 13) | 8% (n = 5) | 10% (−27% to 47%) |
Most respondents indicated routine vital signs were obtained every 4 hours (78%; 95% confidence interval, 67%–87%), with every 8 hours as the second-most common response (14%; 95% CI, 7%–23%). When queried specifically, blood pressure and pulse oximetry measurement tracked closely with default vital sign response; however, 10% of hospitalists reporting every 4-hour routine vitals reported obtaining blood pressures less frequently than the other vital signs (Table 2).
Frequency of Routine Vital Signs
| . | Vitals (95% CI) . | BP (95% CI) . | Pulse Ox (95% CI)a . |
|---|---|---|---|
| Every 4 h | 79% (67%–87%) | 68% (56%–78%) | 77% (64%–85%) |
| Every 6 h | 6% (2%–15%) | 5% (2%–15%) | 4% (1%–11%) |
| Every 8 h | 14% (7%–23%) | 22% (13%–33%) | 14% (7%–23%) |
| Every 12 h, or less | 1% (0%–7%) | 5% (2%–15%) | 4% (1%–11%) |
| . | Vitals (95% CI) . | BP (95% CI) . | Pulse Ox (95% CI)a . |
|---|---|---|---|
| Every 4 h | 79% (67%–87%) | 68% (56%–78%) | 77% (64%–85%) |
| Every 6 h | 6% (2%–15%) | 5% (2%–15%) | 4% (1%–11%) |
| Every 8 h | 14% (7%–23%) | 22% (13%–33%) | 14% (7%–23%) |
| Every 12 h, or less | 1% (0%–7%) | 5% (2%–15%) | 4% (1%–11%) |
BP, blood pressure; pulse ox, pulse oximetry.
One institution measures pulse oximetry continuously (1%).
Absolute differences in means in bed size (δ = 26; 95% CI, −62 to 114), trauma center designation (δ = 2%; 95% CI, −37% to 41%), or CHA membership (δ = 4%, 95% CI, −14% to 22%) between hospitalists reporting default vital sign frequency of 4 hours and those reporting less frequent measurements were not statistically significant.
Twenty-one (28%) survey participants added optional free-text comments or clarifications. Comments were organized into 2 common themes:
intentionally decreasing frequency of routine vital signs over the course of the hospital stay; and
intentionally decreasing frequency of routine vital signs on the basis of age, diagnosis, or other specific criteria.
Representative quotes related to these themes can be found in Table 3.
Representative Examples of Most Frequent Themes in Survey Free-Text Comments
| Theme 1: Intentionally Decreasing Vital Sign Frequency Over the Course of the Hospital Stay . |
|---|
| “At our institution, routine vital signs are q4 h × 24 h and then q shift (12 h) unless otherwise specified.” “q4 [h] vitals for first 24 h, then q8 h thereafter (unless degree of illness requires continuing q4 h).” “The vital signs order is every 4 h × 24 h, then every 8 h.” “We instituted a ‘nighttime vital signs protocol’ a couple of y ago to decrease overnight wake-ups in otherwise stable kids. This has been helpful and well received by families and staff. No negative outcomes to date.” “We do have a sleep initiative where stable patients do not have vitals taken physically at the bedside but remotely.” |
| Theme 2: intentionally decreasing vital sign frequency on the basis of age, diagnosis, or other criteria |
| “Depending on the medical condition, BP is not routinely collected as a routine vital sign in all children <12 mo of age in our institution.” “BP measurement varies by age. Children <2 y of age by default get BPs once every 24 h unless otherwise ordered.” “Our facility has a policy that excludes midnight BP on stable patients <3 y of age to limit disruptions to sleep.” “We do skip vitals overnight over the age of 2 in stable patients. Pulse ox depends on condition. Some need it all the time, some q4 h, and some none of the time.” “We have an order set for ‘reduced vital signs.’ If a patient has 2 or more PEWS scores of 1, vital signs frequency can be decreased to once per shift with no invasive vitals taken when the patient is asleep (temp, BP).” “We often get requests to decrease nighttime checks to promote better sleep in stable patients.” |
| Theme 1: Intentionally Decreasing Vital Sign Frequency Over the Course of the Hospital Stay . |
|---|
| “At our institution, routine vital signs are q4 h × 24 h and then q shift (12 h) unless otherwise specified.” “q4 [h] vitals for first 24 h, then q8 h thereafter (unless degree of illness requires continuing q4 h).” “The vital signs order is every 4 h × 24 h, then every 8 h.” “We instituted a ‘nighttime vital signs protocol’ a couple of y ago to decrease overnight wake-ups in otherwise stable kids. This has been helpful and well received by families and staff. No negative outcomes to date.” “We do have a sleep initiative where stable patients do not have vitals taken physically at the bedside but remotely.” |
| Theme 2: intentionally decreasing vital sign frequency on the basis of age, diagnosis, or other criteria |
| “Depending on the medical condition, BP is not routinely collected as a routine vital sign in all children <12 mo of age in our institution.” “BP measurement varies by age. Children <2 y of age by default get BPs once every 24 h unless otherwise ordered.” “Our facility has a policy that excludes midnight BP on stable patients <3 y of age to limit disruptions to sleep.” “We do skip vitals overnight over the age of 2 in stable patients. Pulse ox depends on condition. Some need it all the time, some q4 h, and some none of the time.” “We have an order set for ‘reduced vital signs.’ If a patient has 2 or more PEWS scores of 1, vital signs frequency can be decreased to once per shift with no invasive vitals taken when the patient is asleep (temp, BP).” “We often get requests to decrease nighttime checks to promote better sleep in stable patients.” |
BP, blood pressure; PEWS, pediatric early warning score; pulse ox, pulse oximetry; q4, every 4; q shift, every shift; temp, temperature.
Discussion
Most hospitalists surveyed reported routine vitals to be those obtained every 4 hours. However, 1 in 5 hospitalists surveyed reported measuring routine vital signs less frequently than every 4 hours, and nearly one-third reported measuring blood pressures less frequently than every 4 hours. Free-text comments highlighted several institutional protocols developed around the notion that patients require less frequent vital sign assessments after the first day of hospitalization, provided no abnormalities are detected. Hospitalists provided varying degrees of specificity regarding how stability was defined, with 1 reporting using an early warning system score to determine which patients qualified for less frequent vital sign measurement. There were also several reports of selectively decreasing the frequency of routine blood pressure measurement in very young children.
The data from our study demonstrate that there is some variability in reported routine vitals, a fact that may inform quality improvement leaders attempting to influence attitudes pertaining to the necessity of frequent monitoring. The knowledge that other institutions have instituted alternative protocols may prompt discussions about the feasibility of changing routine monitoring frequency of select patient populations. Survey respondents highlighted several conditions under which deviation from initial vital sign monitoring practices may occur.
Of the hospitalists who reported routine vital signs as those collected every 4 hours, many indicated that blood pressures were obtained less frequently. Three hospitalists reported that children younger than 1 to 3 years specifically had blood pressures measured less frequently at their institutions. Although blood pressure norms are defined on the basis of auscultatory readings in the arms of upright patients, alternative collection methods (eg, in the legs of supine patients) are commonly employed in the hospital setting to avoid sites of intravenous lines and to avoid awakening or agitating children (particularly young children). These practices may inadvertently falsely elevate or depress blood pressure measurements.12,13 When falsely elevated or depressed blood pressures are obtained, they may result in unnecessary and disruptive additional measurements or interventions and/or they can be dismissed as inaccurate.
In previously published work, Cook et al discusses a quality improvement intervention at their institution specifically targeting the measurement of blood pressure in hospitalized pediatric patients.14 By decreasing the number of patients with blood pressures measured overnight, they increased in-hospital sleep duration for children aged 2 years and older, decreased nighttime disruptions from clinicians by 19%, and saved >$15 000 in estimated cost, without demonstrating an increase in adverse events. Lee et al implemented a sleep-promoting protocol at their institution with only “passive” vital signs (heart rate, respiratory rate, and oximetry) collected at 4 am in select, stable patients.15
The scope of each of these projects was limited. The intervention of Cook et al focused only on reducing the frequency of the midnight and 4 am blood pressure measurement, whereas Lee et al focused only on the 4 am blood pressure and temperature. The data from our report can be used by quality improvement teams to build on the previous work of Cook and Lee et al. Our findings suggest that routine vital signs are not a normative concept across all patient populations in pediatrics and that it is reasonable to consider alternative default vital sign frequencies for hospitalized pediatric patients because others are already doing so. Our report also provides several concrete change ideas for those undertaking quality improvement interventions designed to decrease unnecessary vital signs.
There are several limitations of our study that may affect the generalizability of the results. By utilizing the AAMC database to identify programs, we only captured hospitals affiliated with a medical school. Despite having an above average response rate to our survey, response bias is possible, though we found no differences between respondents and nonrespondents in bed size, trauma center designation, CHA membership, and geographic region. We acknowledge that the vital sign frequency reported by hospitalists in this study has a high risk of failing to correlate with observed practice; however, for the purpose of change management, we intentionally sought to understand self-reported behavior rather than actual practice.
Acknowledgments
We thank Drs Jimmy Beck, Samantha House, and Corrie McDaniel for their feedback on the survey. We also thank Drs Brian Alverson and H. Barrett Fromme for participating in cognitive interviews.
FUNDING: No external funding.
CONFLICT OF INTEREST DISCLAIMER: The authors have indicated they have no conflicts of interest relevant to this article to disclose.
Drs McDaniel and Ralston conceptualized and designed the study, performed the analyses, drafted and edited the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.
Comments