SEARCH

SEARCH BY CITATION

Keywords:

  • Systematic review;
  • prognosis;
  • vital signs;
  • blood gas monitoring;
  • transcutaneous;
  • hospital

Abstract

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Background

Conflicting evidence exists on the effectiveness of routinely measured vital signs on the early detection of increased probability of adverse events.

Purpose

To assess the clinical relevance of routinely measured vital signs in medically and surgically hospitalized patients through a systematic review.

Data Sources

MEDLINE, Embase, Cochrane Central Register of Controlled Trials (CENTRAL), Cumulative Index to Nursing and Allied Health Literature, and Meta-analysen van diagnostisch onderzoek (in Dutch; MEDION) were searched to January 2013.

Study Selection

Prospective studies evaluating routine vital sign measurements of hospitalized patients, in relation to mortality, septic or circulatory shock, intensive care unit admission, bleeding, reoperation, or infection.

Data Extraction

Two reviewers independently assessed potential bias and extracted data to calculate likelihood ratios (LRs) and predictive values.

Data Synthesis

Fifteen studies were performed in medical (n = 7), surgical (n = 4), or combined patient populations (n = 4; totaling 42,565 participants). Only three studies were relatively free from potential bias. For temperature, the positive LR (LR+) ranged from 0 to 9.88 (median 1.78; n = 9 studies); heart rate 0.82 to 6.79 (median 1.51; n = 5 studies); blood pressure 0.72 to 4.7 (median 2.97; n = 4 studies); oxygen saturation 0.65 to 6.35 (median 1.74; n = 2 studies); and respiratory rate 1.27 to 1.89 (n = 3 studies). Overall, three studies reported area under the Receiver Operator Characteristic (ROC) curve (AUC) data, ranging from 0.59 to 0.76. Two studies reported on combined vital signs, in which one study found an LR+ of 47.0, but in the other the AUC was not influenced.

Conclusions

Some discriminative LR+ were found, suggesting the clinical relevance of routine vital sign measurements. However, the subject is poorly studied, and many studies have methodological flaws. Further rigorous research is needed specifically intended to investigate the clinical relevance of routinely measured vital signs.

Clinical Relevance

The results of this research are important for clinical nurses to underpin daily routine practices and clinical decision making.

Doctors and nurses have traditionally been taught that routine monitoring of vital signs is an important way of measuring physiological functioning and determining the probability of clinical deterioration and adverse events (Evans, Hodgkinson, & Berry, 2001; Kammersgaard et al., 2001). Instructions to monitor vital signs are widely found in textbooks, in clinical teaching, and on ward rounds during which patients’ vital sign charts are studied and discussed. Although routine monitoring is daily practice in hospitals, its diagnostic effectiveness has been a point of debate for many years.

Many older studies conclude that measuring vital signs is useful. These studies suggest that changes in vital signs occur hours prior to adverse events and clinical deterioration (Goldhill, White, & Sumner, 1999; Payman, Dampier, & Hawthorn, 1989; Schein, Hazday, Pena, Ruben, & Sprung, 1990). This is in direct contrast to more recent studies, which question the relevance of routine measurements. More recent studies have come to the conclusion that changes in vital signs either do not occur or do not occur early enough to determine the probability of adverse events in general hospital patients (Conen, Leimenstoll, Perruchoud, & Martina, 2006; Vermeulen, Storm-Versloot, Goossens, Speelman, & Legemate, 2005; Zeitz & McCutcheon, 2006).

The prevention and early detection of clinical deterioration and adverse events is currently a major topic in quality assurance programs. Worldwide, several governmental institutes have developed guidelines on the identification of acutely ill medical patients that recommend the use of early warning scores or related systems in which vital sign measurements are combined in an overall score (Institute of Healthcare Improvement, n.d.; Smith, 2011). Implementation of these guidelines has led to a substantial increase in the measuring of vital signs. Recent literature has mainly focused on the accuracy of these early warning models, which provide clinicians a tool for severity assessment (Gao et al., 2007; McGaughey et al., 2007). Knowledge of the positive likelihood ratio (LR+) for the different thresholds of each vital sign within these models is important in order to interpret them.

Purpose

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

This systematic review was initiated to identify and summarize those studies that have examined the clinical relevance of each routinely measured vital sign in detecting adverse events (mortality, septic shock, circulatory shock, admission to the intensive care unit [ICU], bleeding, reoperation, and infection) in medically and surgically hospitalized patients.

Methods

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Data Sources

Search strategy

Systematic and comprehensive searches were developed with a clinical librarian. The electronic databases MEDLINE, Embase, CENTRAL , Cumulative Index to Nursing and Allied Health Literature, and MEDION were searched up to January 2013. Search terms and strategies for each database are described in Addendum 1 (found with the online version of this article).

In addition to these searches, one author contacted experts in the field to identify relevant ongoing publications. However, we did not search abstract books of conference proceedings.

Study selection

We performed a three-phase selection process. The search identified 15,947 citations of potential relevance. In the first phase, two reviewers independently checked a 25% random sample of the abstracts of these references with broad inclusion criteria (Addendum 2; found with the online version of this article). If there was disagreement, the abstracts were discussed. Since agreement was more than 95%, one reviewer continued with the remaining abstracts.

Two reviewers independently performed a second selection of the remaining 451 abstracts with narrower inclusion criteria for mortality, septic shock, circulatory shock, admission to the ICU, bleeding, reoperation, and infection outcomes. It also had to be possible to create 2 × 2 tables from data reported in the abstract. Any disagreement was resolved through discussion.

The overall selection revealed 61 potentially relevant studies for which the full text was independently read in the third phase by a team member who created 2 × 2 tables of extracted data as a final inclusion criterion. In cases of disagreement, the final decision regarding inclusion was made by a third reviewer.

Criteria for Inclusion

Types of studies

Inclusion criteria were any prosp-ective study that evaluated routine measurements of vital signs in an original hospitalized patient series, had been published as full text, and written in English, French, German, Dutch, or Spanish. If a number of studies had been published on the same series, the most complete study on outcomes was used.

Types of participants

Inclusion criteria were surgical and medical patients at least 18 years of age and admitted to general hospital wards. Patients admitted to specialized wards, such as the ICU, cardiac care unit, neurology, and cardiology were excluded, as were patients only measuring vital signs within the first 24 hr after admission for medical treatment or surgical intervention.

Types of interventions

Inclusion criteria were studies measuring vital signs (temperature, heart rate, blood pressure, oxygen saturation, and respiratory rate) on a routine basis. The vital signs had to be compared to a reference standard, which could be a single test or a combination of different tests or could be defined as the presence or absence of an adverse event (Fletcher & Fletcher, 2005).

Types of outcome measures

Adverse events of interest were mortality, septic or circulatory shock, admission to the ICU, bleeding, reoperation, or infection. Information on vital signs in relation to adverse events had to be reported as sensitivity, specificity, predictive values, LRs, and area under the ROC curve (AUC); alternatively, the 2 × 2 table could be calculated from information given in the report. These clinical relevance parameters reflect the increased probability of having outcomes of interest.

Data Extraction

Two reviewers independently extracted data using a data extraction sheet. To describe the included studies in sufficient detail and to identify clinical heterogeneity, we extracted the following data: study characteristics (study design, year of publication, specialty , and country of origin); patient demographics (age, gender, type of disease, reason for admission, and comorbidity); vital sign measurements (frequencies, threshold used, moment of measurement); adverse events, with their definition and prevalence (mortality, septic or circulatory shock, admission to ICU, bleeding, reoperation, and infection); reference tests; and the above-mentioned types of outcome measures or raw study data to calculate clinical relevance parameters.

Quality Assessment

A valid checklist for prognostic studies has been lacking until now; however, six domains with potentially useful criteria have been described by Hayden, Cote, and Bombardier (2006) and Minne, Abu-Hanna, and De Jonge (2008). Two of these domains are relevant to our study and are congruent with four relevant quality items of the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool (Addendum 3, found with the online version of this article; Whiting, Rutjes, Reitsma, Bossuyt, & Kleijnen, 2003). These four criteria comprise a defined representative patient spectrum, a defined outcome such as presence or absence of an adverse event, independent measurements, and blind outcome assessment. The methodological quality was assessed independently by two reviewers and is summarized in Table S1 (found with the online version of this article).

Analysis

From each study, 2 × 2 tables were extracted for each routinely measured vital sign and for each threshold in order to calculate clinical relevance parameters: sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), LR+, and negative likelihood ratio (LR-). Additionally, a 95% confidence interval (95% CI) was calculated for each. The area under the receiver operator characteristic curve (AUC) data was extracted from studies when it was reported. The AUC represents the overall accuracy of a test and is calculated from plotting of the Sensitivity against “1-Specificity” to a single value representing the expected performance. An AUC of 100% represent perfect discrimination, whereas an AUC of 50% represent no discrimination.

Results

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Characteristics of Included Studies

Fifteen studies (Arnell, De Virgilio, Chang, Bongard, & Stabile, 1996; Barbier, Dieu, & Dumora, 1986; Chalmers, Singanayagam, & Hill, 2008; Conen et al., 2006; Gomez et al., 2006; Gonen, Turan, Ozturk, & Ozkardes, 2008; Hoogewerf, Oosterheert, Hak, Hoepelman, & Bonten, 2006; Kline, Hernandez-Nino, Rose, Norton, & Camargo, 2006; Lighthall, Markar, & Hsiung, 2009; Madan, Stoecklein, Ternovits, Tichansky, & Phillips, 2007; Mato et al., 2009, 2010; Payman et al., 1989; Smith et al., 2012; Vermeulen et al., 2005) met the inclusion criteria(see Figure 1),found with the online version of this article). All were published between 1986 and 2012, with a total of 42,565 participants. Details are summarized in Table 1.

Table 1. Characteristics of Included Studies
     ThresholdMean yearsSD   Reason
AuthorYearCountryOutcomeVital signsvital signsof age;or rangeNMaleSpecialtyfor admission
Note
  1. T = temperature, HR = heart rate, BP = blood pressure, SpO2 = peripheral oxygen saturation, RR = respiratory rate. I = intervention group, C = comparison group, SD = standard deviation, NA = not applicable.

  2. a

    Combined outcomes;

  3. b

    Combined vital signs;

  4. c

    Single and combined vital signs;

  5. d

    single and combined outcomes.

Smith et al.2012UKMortalitySpO2Yes64.4SD 20.537,59317,779MedicalAcute medical admission
Chalmers et al.2008UKMortality; ICU admissionBPYes66Range 50–781,007500MedicalCommunity-acquired pneumonia
Hoogewerf et al.2006NetherlandsMortality, ICU admissionaT, HR, BP, RRYesI: 67.9; C: 69.2 260180MedicalCommunity-acquired pneumonia
Kline et al.2006USAMortality, ICU admission, circulatory shockaSpO2Yes53SD 1720086CombinedPulmonary embolism
Conen et al.2006SwitzerlandICU admission, bleeding, infectionaBPYes64SD 17639322CombinedAll cause
Lighthall et al.2009USAMortality; ICU admission cardiopulmonary arrestdHR, BP, RR, SpO2bYes66SD 17.51,08996%CombinedAll cause
Arnell et al.1996USAICU admissionHRYes37Range 17–8310212CombinedGallstone pancreatitis
Gomez et al.2006ArgentinaICU admissionT, HR, BP, RRYes48Range 16–83167NAMedicalNeutropenia with fever
Gonen et al.2008TurkeySeptic shockTYesI: 47;SD 19.361NAMedicalKidney stones
      C: 46.5SD 14.8    
Mato et al.2009USASeptic shockT, HR, RRcYesI: 54; 547NAMedicalHematologic
      C: 51    malignancies
Mato et al.2010USASeptic shockT, HR, RRYesI: 54;     
      C: 51 230126MedicalHematologic malignancies
Barbier et al.1986FranceInfectionTYes63.5Range 40–8010035SurgicalHip arthroplasty
Payman et al.1989UKInfectionTYes56.9SD 19.64132SurgicalElective surgery
Vermeulen et al.2005NetherlandsInfectionTYes55.3SD 16.1284134SurgicalAll elective surgery patients
Madan et al.2007USAReoperationT, HRYes39Range 17–66245NASurgicalLaparoscopic gastric bypass

Data Synthesis

Studies were too heterogeneous to pool data. Therefore, the results are presented as clinical relevance parameters according to each vital sign in relation to the outcomes of interest.

Clinical Relevance Parameters of Vital Signs

Temperature (Table 2A)

Nine studies (Barbier et al., 1986; Gomez et al., 2006; Gonen et al., 2008; Hoogewerf et al., 2006; Madan et al., 2007; Mato et al., 2009, 2010; Payman et al., 1989; Vermeulen et al., 2005) reported on temperature, for which the most common outcomes were infection or septic shock. LR+ ranged from 0 to 9.88 (median 1.78). The highest LR+ was found for reoperation outcomes, with a threshold of > 38.6°C. The accompanying PPV increased from 3.3% to 25% (Madan et al., 2007). The lowest LR+ was found in the primary diagnostic design study (Vermeulen et al., 2005) for infection outcomes with a threshold ≥ 39°C. The accompanying PPV decreased from 6.7% to 0%. One study (Mato et al., 2009) reported AUC data, which ranged from 59 to 61. There were no distinctive differences between medical and surgical study populations.

Table 2. Clinical Relevance Parameters in Relation to Defined Outcomes
      Sens %Spec % (PPV %NPV %LR+LR-AUC %
AuthorSpecialtyThresholdNOutcomePrev %(95% CI)95% CI)(95% CI)(95% CI)(95% CI)(95% CI)(95% CI)
A: Temperature (°C)
Note
  1. Prev = prevalence; Sens = sensitivity; Spec = specificity; PPV = positive predictive value; NPV = negative predictive value; LR+ = positive likelihood ratio; LR- = negative likelihood ratio; AUC = area under the curve; CI = confidence interval; ICU = intensive care unit; NA = not applicable; Temp = body temperature; HR bpm = heart rate in beats per minute; BP = blood pressure; SpO2 = peripheral oxygen saturation; RR/min = respiratory rate per minute; Syst = systolic; Diast = diastolic.

  2. a

    Measurements;

  3. b

    Combined outcome;

  4. c

    HR bpm < 40 or > 110, systolic BP mmHg < 90, SpO2 % < 90, RR/min < 8 or > 26;

  5. d

    Difference in mmHg from baseline blood pressure.

Hoogewerf et al. (2006)Medical< 35 or > 40260Mortality, ICUb30.814 (6–21)87 (82–92)31 (16–47)69 (63–75)1.03 (0.53–2.00)1.0 (0.90–1.11)NA
Gomez et al. (2006)Medical> 39167ICU40.128 (18–39)77 (71–84)33 (21–45)73 (67–80)1.24 (0.78–1.99)0.93 (0.78–1.10)NA
Gonen et al. (2008)Medical≥ 3861Septic shock1.610085 (76–94)10 (0–29)1006.67 (3.65–12.18)0.000NA
Mato et al. (2009)Medical< 36 or > 38547Septic shock8.4NANANANANANA61 (54–68)
Mato et al. (2009)Medical≥ 37.9547Septic shock8.4NANANANANANA59 (53–65)
Mato et al. (2010)Medical<36 or > 38547Septic shock8.410044 (39–48)14 (10–18)1001.78 (1.65–1.92)0.00NA
Barbier et al. (1986)Surgical≥ 38100Infection3.01008 (3–14)3 (0–7)1001.09 (1.03–1.16)0.00NA
Payman et al. (1989)SurgicalTemp oral ≥ 37.441Infection29.358 (30–86)72 (56–89)47 (21–72)81 (66–96)2.12 (0.99–4.52)0.58 (0.28–1.17)NA
Payman et al. (1989)SurgicalTemp aural ≥ 37.841Infection29.358 (30–86)69 (52–86)44 (19–68)80 (64–96)1.88 (0.91–3.87)0.60 (0.30–1.23)NA
Vermeulen et al. (2005)Surgical1 sign ≥ 38284Infection6.737 (15–59)80 (75–85)12 (4–20)95 (92–98)1.81 (0.96–3.41)0.79 (0.56–1.12)NA
Vermeulen et al. (2005)Surgical≥ 38.5284Infection6.75 (0–15)93 (90–96)5 (0–15)93 (90–96)0.73 (0.10–5.19)1.02 (0.91–1.14)NA
Vermeulen et al. (2005)Surgical≥ 39284Infection6.7098 (97–100)093 (90–96)0.001.02 (1.00–1.04)NA
Vermeulen et al. (2005)Surgical2 signs ≥ 38.0284Infection6.711 (0–24)92 (88–95)8 (0–19)94 (91–97)1.27 (0.32–4.99)0.98 (0.83–1.14)NA
Vermeulen et al. (2005)SurgicalAll signs ≥ 38.02,282aInfection6.76 (3–9)92 (91–94)8 (4–12)90 (89–91)0.81 (0.48–1.38)1.02 (0.98–1.05)NA
Vermeulen et al. (2005)Surgical≥ 38.52,282aInfection6.70 (0–1)98 (98–99)3 (0–7)90 (89–91)0.23 (0.03–1.69)1.02 (1.00–1.03)NA
Vermeulen et al. (2005)Surgical≥ 39.02,282aInfection6.70100090 (89–91)0.001.00NA
Madan et al. (2007)Surgical> 38.6245Reoperation3.325 (0–55)98 (96–100)25 (0–55)98 (96–100)9.88 (2.05–41.6)0.77 (0.52–1.15)NA
B: Heart rate (bpm)
Hoogewerf et al. (2006)Medical> 125260Mortality, ICUb30.826 (17–36)78 (72–84)35 (23–47)71 (64–77)1.21 (0.76–1.92)0.94 (0.81–1.10)NA
Arnell et al. (1996)Combined≥ 100102ICU6.986 (60–100)87 (81–94)33 (12–55)99 (97–100)6.79 (3.69–12.48)0.16 (0.03–1.01)NA
Gomez et al. (2006)Medical> 120167ICU40.127 (16–38)81 (75–87)35 (22–48)74 (68–80)1.81 (0.84–2.30)0.79 (0.77–1.07)NA
Mato et al. (2009)Medical> 90547Septic shock8.4NANANANANANA64 (57–71)
Mato et al. (2009)Medical≥ 99547Septic shock8.4NANANANANANA71 (63–78)
Madan et al. (2007)Surgical> 100245Reoperation3.313 (0–35)85 (80–89)3 (0–8)97 (94–99)0.82 (0.13–5.28)1.03 (0.79–1.35)NA
C: Blood pressure (mmHg)
Chalmers et al. (2008)MedicalSyst < 901,007Mortality9.641 (31–51)90 (88–92)31 (23–38)93 (92–95)4.07 (3.00–5.54)0.66 (0.55–0.77)70 (67–74)
Chalmers et al. (2008)MedicalDiast ≤ 601,007Mortality9.654 (44–64)65 (62–68)14 (10–18)93 (91–95)1.52 (1.24–1.87)0.72 (0.58–0.89)59 (56–62)
Conen et al. (2006)CombinedSyst > 10d639Mortality, ICU, bleeding, infectionb19.141 (32–50)59 (55–63)18 (14–23)81 (77–85)0.99 (0.78–1.26)1.01 (0.86–1.19)NA
Conen et al. (2006)CombinedSyst > 20d639Mortality, ICU, bleeding, infectionb19.110 (5–16)86 (83–89)14 (7–22)81 (77–84)0.72 (0.41–1.28)1.05 (0.98–1.12)NA
Conen et al. (2006)CombinedDiast > 10d639Mortality, ICU, bleeding, infectionb19.124 (16–31)76 (72–79)18 (12–24)81 (78–85)0.97 (0.68–1.39)1.01 (0.90–1.13)NA
Conen et al. (2006)CombinedDiast > 20d639Mortality, ICU, bleeding, infectionb19.15 (1–9)95 (93–97)19 (5–33)82 (78–97)1.05 (0.44–2.51)1.0 (0.95–1.04)NA
Hoogewerf et al. (2006)MedicalSyst < 90260Mortality, ICUb30.84 (0–8)98 (96–100)43 (6–80)70 (63–75)1.69 (0.39–7.37)0.98 (0.94–1.03)NA
Chalmers et al. (2008)MedicalSyst < 901,007ICU10.245 (35–54)91 (89–92)35 (27–43)93 (92–95)4.70 (3.50–6.30)0.61 (0.51–0.73)70 (67–73)
Chalmers et al. (2008)MedicalDiast ≤ 601,007ICU10.265 (56–74)66 (63–69)18 (14–22)94 (93–96)1.93 (1.62–2.27)0.52 (0.40–0.69)68 (65–72)
Gomez et al. (2006)MedicalSyst < 90167ICU40.122 (12–32)95 (91–98)63 (43–82)76 (70–81)4.25 (1.96–9.25)0.82 (0.72–0.94)NA
D: Peripheral oxygen saturation (SpO2 %)
Kline et al. (2006)Combined< 95200Mortality, ICU, circulatory shockbUnknown65 (52–78)67 (59–75)43 (32–55)83 (76–90)1.96 (1.42–2.69)0.53 (0.36–0.78)NA
Smith et al. (2012)Medical≤ 8837,593Mortality5.36 (5–7)99 (99–99)26 (22–30)95 (95–95)6.35 (5.17–7.80)0.95 (0.94–0.96)NA
Smith et al. (2012)Medical8937,593Mortality5.32 (1–2)99 (99–100)19 (13–26)95 (95–95)4.31 (2.90–6.41)0.99 (0.98–0.99)NA
Smith et al. (2012)Medical9037,593Mortality5.33 (2–4)99 (99–99)16 (12–20)95 (95–95)3.44 (2.62–4.53)0.98 (0.97–0.99)NA
Smith et al. (2012)Medical9137,593Mortality5.32 (2–3)99 (99–99)11 (8–14)95 (95–95)2.27 (1.69–3.06)0.99 (0.98–0.99)NA
Smith et al. (2012)Medical9237,593Mortality5.35 (4–6)98 (98–98)11 (9–13)95 (95–95)2.23 (1.80–2.75)0.97 (0.96–0.98)NA
Smith et al. (2012)Medical9337,593Mortality5.38 (6–9)96 (96–97)10 (9–12)95 (95–95)2.07 (1.75–2.43)0.96 (0.95–0.97)NA
Smith et al. (2012)Medical9437,593Mortality5.310 (9–11)93 (93–94)8 (7–9)95 (95–95)1.51 (1.31–1.73)0.96 (0.95–0.98)NA
Smith et al. (2012)Medical9537,593Mortality5.313 (12–15)88 (87–88)6 (5–6)95 (95–95)1.08 (0.96–1.21)0.99 (0.97–1.01)NA
Smith et al. (2012)Medical9637,593Mortality5.315 (14–17)82 (82–82)5 (4–5)95 (94–95)0.84 (0.76–0.94)1.03 (1.01–1.05)NA
Smith et al. (2012)Medical9737,593Mortality5.312 (11–14)82 (81–82)4 (3–4)94 (94–95)0.68 (0.60–0.77)1.07 (1.05–1.09)NA
Smith et al. (2012)Medical9837,593Mortality5.312 (10–13)82 (82–82)3 (3–4)94 (94–95)0.65 (0.57–0.73)1.08 (1.06–1.10)NA
Smith et al. (2012)Medical9937,593Mortality5.38 (7–9)89 (89–89)4 (3–4)95 (94–95)0.71 (0.61–0.83)1.04 (1.02–1.05)NA
Smith et al. (2012)Medical10037,593Mortality5.34 (4–5)93 (93–94)4 (3–4)95 (94–95)0.67 (0.54–0.82)1.02 (1.01–1.03)NA
Smith et al. (2012)Medical< 9537,593Mortality5.335 (33–37)84 (84–85)11 (10–12)96 (96–96)2.25 (2.11–2.40)0.77 (0.75–0.80)NA
E: Respiratory rate (/min)
Hoogewerf et al. (2006)Medical> 30260Mortality, ICUb30.839 (28–49)69 (63–76)36 (26–46)72 (65–79)1.27 (0.89–1.81)0.88 (0.72–1.08)NA
Gomez et al. (2006)Medical> 24167ICU40.125 (15–36)87 (84–92)43 (27–58)75 (69–81)1.89 (1.08–3.30)0.86 (0.74–1.00)NA
Mato et al. (2009)Medical> 20547Septic shock8.4NANANANANANA59 (53–65)
F: Combined vital signs
Lighthall et al. (2009)Combined1 abnormal signc1,089Mortality, ICU, cardiac arrestb7.672 (61–81)89 (87–91)35 (28–42)97 (96–98)6.61 (5.30–8.24)0.31 (0.22–0.44)NA
Lighthall et al. (2009)Combined1 abnormal signc1,089Mortality 30 days4.259 (43–73)86 (84–88)16 (11–22)98 (97–99)4.28 (3.22–5.70)0.48 (0.34–0.68)NA
Lighthall et al. (2009)Combined1 abnormal signc1,089ICU3.589 (75–97)87 (85–89)20 (14–27)99 (98–100)6.91 (5.71–8.37)0.12 (0.05–0.31)NA
Lighthall et al. (2009)Combined≥ 2 abnormal signsc1,089Mortality, ICU, cardiac arrestb7.628 (18–39)99 (99–99)78 (57–91)NA47.00NANA
Lighthall et al. (2009)Combined≥ 2 abnormal signsc1,089Mortality 30 days4.226 (14–41)99 (97–99)44 (25–65)NA18.10NANA
Lighthall et al. (2009)Combined≥ 2 abnormal signsc1,089ICU3.537 (22–54)99 (98–99)52 (32–71)NA29.80NANA
Mato et al. (2009)MedicalTemp °C < 36 or >38 HR bpm > 90 RR/min > 20547Septic shock8.4NANANANANANA73 (64–80)
Mato et al. (2009)MedicalHR bpm > 90 RR/min > 20547Septic shock8.4NANANANANANA75 (67–82)
Mato et al. (2009)MedicalTemp °C < 36 or > 38 RR/min > 20547Septic shock8.4NANANANANANA68 (60–75)
Mato et al. (2009)MedicalTemp °C ≥ 37.9, HR ≥ 99, RR/min > 20547Septic shock8.4NANANANANANA76 (68–84)
Mato et al. (2009)MedicalHR bpm ≥ 99 RR/min > 20547Septic shock8.4NANANANANANA74 (66–81)
Mato et al. (2009)MedicalTemp °C ≥ 37.9, RR/min > 20547Septic shock8.4NANANANANANA66 (59–73)
Mato et al. (2010)MedicalHR bpm > 90 RR/min > 20547Septic shock8.4NANANANANANA68 (61–77)
Heart rate (Table 2B)

Five studies (Arnell et al., 1996; Gomez et al., 2006; Hoogewerf et al., 2006; Madan et al., 2007; Mato et al., 2009) reported on heart rate, for which the most common outcome was ICU admission. LR+ ranged from 0.79 to 6.79 (median 1.51). The highest LR+ was found for ICU admission outcomes, with a threshold of ≥ 100 bpm (Arnell et al., 1996). The accompanying PPV increased from 6.9% to 33%. It was notable that using the same threshold, a LR+ of 0.82 was found for reoperation outcomes (Madan et al., 2007). One study (Mato et al., 2009) reported AUC data, which ranged from 64 to 71, dependent on which threshold was used.

One study (Madan et al., 2007) reported on a surgical population. There were no distinctive differences between medical, surgical, and unknown study populations.

Blood pressure (Table 2C)

Four studies (Chalmers et al., 2008; Conen et al., 2006; Gomez et al., 2006; Hoogewerf et al., 2006) reported on blood pressure, in which outcomes were mortality and ICU admission. LR+ ranged from 0.72 to 4.70 (median 2.97). The highest LR+ was found for ICU admissions, with a threshold for systolic blood pressure of < 90 mmHg (Chalmers et al., 2008). The accompanying PPV increased from 10.2% to 35%. The same study (Chalmers et al., 2008) reported AUC data, which ranged from 59 to 70. It was notable that a difference of 20 mmHg from baseline systolic blood pressure showed a LR+ of 0.726 (Conen et al.).

No study specifically reported on a surgical population. There were no distinctive differences between medical and combined study populations.

Peripheral oxygen saturation (Table 2D)

Two studies (Kline et al., 2006; Smith et al., 2012) reported on oxygen saturation, in which mortality and a combined outcome was measured. LR+ ranged from 0.65 to 6.35 (median 1.74). The highest LR+ was found for mortality, with a threshold for peripheral oxygen saturation (SpO2) of ≤ 88% (Smith et al., 2012). The accompanying PPV increased from 5.3% to 26%. No study specifically reported on a surgical population. There were no distinctive differences between medical and combined study populations.

Respiratory rate (Table 2E)

Three studies (Gomez et al., 2006; Hoogewerf et al., 2006; Mato et al., 2009) reported on respiratory rate, with different outcomes. LR+ ranged from 1.27 to 1.89. The highest LR+ was found for ICU admission, with a threshold of > 24/min (Gomez et al., 2006). The accompanying PPV increased from 40.1% to 43%. One study (Mato et al., 2009) reported an AUC of 59. No study specifically reported on a surgical population.

Single and combinations of deteriorated vital signs (Table 2F)

One study (Lighthall et al., 2009) reported on abnormal single or combinations of vital signs, with outcomes of a combination of mortality and admission to the ICU. For one abnormal sign, LR+ ranged from 4.28 to 6.91. For two abnormal signs, LR+ ranged from 18.10 to 47.00. It was notable that although the accompanying PPV was increasing, the sensitivity was decreasing. In other words, of those patients who had an adverse event, 28% had two abnormal signs.

Two studies (Mato et al., 2009, 2010) reported on the relationship between multiple combinations of deteriorated temperature, heart rate, respiratory rate, and septic shock. The AUC ranged from 66 to 76. There was no difference in discriminating power between two and three deteriorated vital signs. No study specifically reported on a surgical population.

Discussion

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Measurement of vital signs is a widely used and accepted routine in hospitalized patients. However, our review revealed that this common practice is poorly studied. Only 14 observational and one diagnostic study were identified, from which only mediocre evidence could be drawn on clinical relevance underpinning this daily practice, especially since most studies were designed for purposes other than our primary study objective and were thus not free of potential bias. In general, all LR+ were low, but some interesting discriminative LR+ for single or combined vital signs were found.

This is illustrated by the study of Chalmers et al. (2008). They reported a promising discriminative LR+ of 4.07 for the single vital sign systolic blood pressure < 90 mmHg. The pre-test probability of 10% rose to a post-test probability of 31%, a change of 21%. However, even when vital signs deviate from the normal value and a relative discriminative LR+ is found, post-test probability remains rather low and the additional value to the clinician is doubtful.

Mato et al. (2009) reported on combined vital signs (temperature, heart rate, and respiratory rate). The AUC ranged from 66 to 76, which was slightly higher than their reports of single vital signs (AUC range 59–71). Lighthall et al. (2009) showed that when two or more abnormal vital signs were present simultaneously, a remarkably high LR+ of 47 was accompanied by a PPV of 78%. However, there is a high proportion of false negative rates; for example, of patients having an adverse event, only 28% had two or more abnormal vital signs.

As well as the poor post-test probabilities and false-negative rates, it is worth discussing the thresholds used and generalizability. The clinical relevance of discriminative LRs found in some of the studies included are questionable, since thresholds of vital signs used are extreme: systolic blood pressure < 90 mmHg (Chalmers et al., 2008; Gomez et al., 2006; Hoogewerf et al., 2006; Lighthall et al., 2009), oxygen saturation < 90% (Lighthall et al., 2009), and respiratory rate < 8 or ≥ 26/min (Lighthall et al., 2009). Patients with these extremes generally have easily identifiable clinical signs of deterioration for doctors and nurses with trained assessment skills or clinical judgment. Furthermore, in the study of Goldhill, McNarry, Mandersloot, and McGinley (2005), which was conducted in patients seen by an intensive care outreach service, similar discriminative LR+ were found when using the same extreme thresholds of vital signs. They also reported on less extreme thresholds, showing that differences between pre-test and post-test probability vanished.

For daily practice it is important to differentiate between thresholds. In our review, only three studies provided results for different thresholds (Mato et al., 2009; Smith et al., 2012; Vermeulen et al., 2005). Mato et al. (2009) used thresholds for heart rate > 90/min or ≥ 99/min, with no significant differences in the AUC. Vermeulen et al. (2005) conducted a diagnostic study and reported on different thresholds of body temperature measurement (BTM) in relation to infection. Results show that BTM is of limited value in the early detection or exclusion of an infection, and the false-negative rate was rather high. Smith et al. (2012) reported each SpO2 value in relation to mortality and showed results identical to those in the study of Goldhill et al. (2005), in which differences between pre-test and post-test probability vanished. This large study, comprising 37,593 medical patients, can be used to interpret different thresholds of oxygen saturation. For example, when using the same threshold as Kline et al. (2006); oxygen saturation < 95), the LR+ was almost equal in both populations. This suggests LR+ is independent of the pre-test probability for this threshold. Furthermore, Smith et al. (2012) showed an NPV of 95% for all thresholds (whether the SpO2 is 100% or lower than 88%), suggesting that in 5% of the SpO2 measurements patients cannot be ruled out to be not at risk. Therefore, one cannot identify patients at risk with routine SpO2 measurements, and the NPV is not informative.

The clinical relevance or generalizability of some studies can be questioned, since specific groups of patients (e.g., community-acquired pneumonia [Chalmers et al., 2008; Hoogewerf et al., 2006], and neutropenia with fever [Gomez et al., 2006]) with high pre-test probability of mortality and ICU admission were studied. Although Chalmers et al. (2008) found some moderate differences from pre-test to post-test, two other studies found none (Gomez et al., 2006; Hoogewerf et al., 2006). The same contradiction can be seen in excluded studies for this review: Goldhill et al. (2005) showed that in patients seen by an intensive care outreach service; an increasing number of deviating vital signs was associated with higher hospital mortality. In contrast, Pedersen, Moller, and Hovhannisyan (2009) found that early detection of hypoxemia in perioperative patients did not reduce either transfer to ICU or mortality. Thus, results are contradictory and can be due to differences in pre-test probability.

Our review demonstrates that there is still a lack of well-designed diagnostic and large observational studies specifically intended to investigate the clinical relevance of routine measurements for patients admitted to general hospital wards. Also, the definitions of “routine measurements” or “nonroutine measurements” are open for debate. This is in line with findings from other literature reviews of vital sign measurements, which conclude that there is a lack of explicit knowledge based on quantitative research (Evans et al., 2001; Lockwood, Conroy-Hiller, & Page, 2004; Pedersen et al., 2009). This suggests that much of the current practice of routinely measuring vital signs in general hospitalized patients (as well as the accuracy, frequency, and usefulness for detecting clinically relevant outcomes) is based on tradition and not yet on evidence from research.

Despite the lack of evidence, the monitoring of vital signs, or models mainly based on vital signs, currently receives a great deal of attention as part of quality and safety programs such as the Survival Sepsis Campaign and campaigns to detect critically ill patients (VMS Veiligheids programma, 2008, 2009). Although observational studies show a relationship between outcomes and the number of patients with deviated vital signs, diagnostic studies can reveal the predictive value of vital signs. Analyzing prospectively sampled large datasets of routinely collected vital signs, as Smith et al. (2012) did, has the potential to determine accuracy regarding prediction of adverse events in hospitalized patients.

Conclusions

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

In this review, several discriminative positive LRs were found, suggesting the clinical relevance of routine vital sign measurements. However, these results must be interpreted with caution because the number of studies was limited, patient groups studied varied largely as to pre-test probability for adverse events, and almost all studies had methodological flaws. The daily routine of measuring vital signs in hospitalized patients can therefore still be questioned. We challenge researchers in the field of vital signs and early warning scores to consider designing and performing diagnostic studies or studies using prospectively collected large datasets. With the present-day popularity of models predicting patients’ adverse events and critical illness, it is important to develop further evidence of the contribution of each routinely measured vital sign.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Additional contributions: Faridi S. van Etten-Jamaludin, clinical librarian at the University of Amsterdam, conducted the literature searches. Piet Bakker, MD, PhD, director of the Department of Quality Assurance & Process Innovation, critically reviewed the manuscript.

Clinical Resources

References

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information
  • Arnell, T. D., De Virgilio, C., Chang, L., Bongard, F., & Stabile, B. E. (1996). Admission factors can predict the need for ICU monitoring in gallstone pancreatitis. American Surgeon, 62, 815819.
  • Barbier, F., Dieu, P., & Dumora, D. (1986). The role of temperature-curve studies in postoperative management of total hip-prosthesis patients. Bordeaux Medical, 19, 6570.
  • Chalmers, J. D., Singanayagam, A., & Hill, A. T. (2008). Systolic blood pressure is superior to other haemodynamic predictors of outcome in community acquired pneumonia. Thorax, 63, 698702.
  • Conen, D., Leimenstoll, B. M., Perruchoud, A. P., & Martina, B. (2006). Routine blood pressure measurements do not predict adverse events in hospitalized patients. American Journal of Medicine, 119(1), (70), e1722.
  • Evans, D., Hodgkinson, B., & Berry, J. (2001). Vital signs in hospital patients: A systematic review. International Journal of Nursing Studies, 38, 643650.
  • Fletcher, R. H., & Fletcher, S. W. (2005). Clinical epidemiology: The essentials. Philadelphia: Lippincott Williams & Wilkins.
  • Gao, H., McDonnell, A., Harrison, D. A., Moore, T., Adam, S., Daly, K., & Harvey, S. (2007). Systematic review and evaluation of physiological track and trigger warning systems for identifying at-risk patients on the ward. Intensive Care Medicine, 33, 667679.
  • Goldhill, D. R., McNarry, A. F., Mandersloot, G., & McGinley, A. (2005). A physiologically-based early warning score for ward patients: The association between score and outcome. Anaesthesia, 60, 547553.
  • Goldhill, D. R., White, S. A., & Sumner, A. (1999). Physiological values and procedures in the 24 h before ICU admission from the ward. Anaesthesia, 54, 529534.
  • Gomez, R. C., Rivero, M., Hugo, K., Novillo, A., Marta, L. M., Recondo, G., & Milberg, M. (2006). [Prognostic risk factors for serious complications in an inpatient population with neutropenia at the onset of a febrile episode]. Medicina (B Aires), 66, 385391.
  • Gonen, M., Turan, H., Ozturk, B., & Ozkardes, H. (2008). Factors affecting fever following percutaneous nephrolithotomy: A prospective clinical study. Journal of Endourology, 22, 21352138.
  • Hayden, J. A., Cote, P., & Bombardier, C. (2006). Evaluation of the quality of prognosis studies in systematic reviews. Annals of Internal Medicine, 144, 427437.
  • Hoogewerf, M., Oosterheert, J. J., Hak, E., Hoepelman, I. M., & Bonten, M. J. M. (2006). Prognostic factors for early clinical failure in patients with severe community-acquired pneumonia. Clinical Microbiology and Infection, 12, 10971104.
  • Institute of Healthcare Improvement (n.d.). Protecting 5 million lives from harm. Retrieved October 27, 2012, from http://www.ihi.org/offerings/Initiatives/PastStrategicInitiatives/5MillionLivesCampaign/Pages/default.aspx
  • Kammersgaard, L. P., Jorgensen, H. S., Reith, J., Nakayama, H., Houth, J. G., Weber, U. J., & Olsen, T. S. (2001). Early infection and prognosis after acute stroke: The Copenhagen Stroke Study. Journal of Stroke and Cerebrovascular Diseases, 10, 217221.
  • Kline, J. A., Hernandez-Nino, J., Rose, G. A., Norton, H. J., & Camargo, C. A., Jr. (2006). Surrogate markers for adverse outcomes in normotensive patients with pulmonary embolism. Critical Care Medicine, 34, 27732780.
  • Lighthall, G. K., Markar, S., & Hsiung, R. (2009). Abnormal vital signs are associated with an increased risk for critical events in US veteran inpatients. Resuscitation, 80, 12641269.
  • Lockwood, C., Conroy-Hiller, T., & Page, T. (2004). Vital signs. JBI Reports, 2, 207230.
  • Madan, A. K., Stoecklein, H. H., Ternovits, C. A., Tichansky, D. S., & Phillips, J. C. (2007). Predictive value of upper gastrointestinal studies versus clinical signs for gastrointestinal leaks after laparoscopic gastric bypass. Surgical Endoscopy, 21, 194196.
  • Mato, A. R., Fuchs, B. D., Heitjan, D. F., Mick, R., Halpern, S. D., Shah, P. D., ... Luger, S. M. (2009). Utility of the systemic inflammatory response syndrome (SIRS) criteria in predicting the onset of septic shock in hospitalized patients with hematologic malignancies. Cancer Biology and Therapy, 8, 10951100.
  • Mato, A. R., Luger, S. M., Heitjan, D. F., Mikkelsen, M. E., Olson, E., Ujjani, C., ... Fuchs, B. D. (2010). Elevation in serum lactate at the time of febrile neutropenia (FN) in hemodynamically-stable patients with hematologic malignancies (HM) is associated with the development of septic shock within 48 hours. Cancer Biology and Therapy, 9, 585589.
  • McGaughey, J., Alderdice, F., Fowler, R., Kapila, A., Mayhew, A., & Moutray, M. (2007). Outreach and early warning systems (EWS) for the prevention of intensive care admission and death of critically ill adult patients on general hospital wards. Cochrane Database of Systematic Reviews, Article No. CD005529.
  • Minne, L., Abu-Hanna, A., & De Jonge, E. (2008). Evaluation of SOFA-based models for predicting mortality in the ICU: A systematic review. Critical Care, 12(6), R161. doi:10.1186/cc7160.
  • Payman, B. C., Dampier, S. E., & Hawthorn, P. J. (1989). Postoperative temperature and infection in patients undergoing general surgery. Journal of Advanced Nursing, 14, 198202.
  • Pedersen, T., Moller, A. M., & Hovhannisyan, K. (2009). Pulse oximetry for perioperative monitoring. Cochrane Database of Systematic Reviews, Article No. CD002013.
  • Schein, R. M., Hazday, N., Pena, M., Ruben, B. H., & Sprung, C. L. (1990). Clinical antecedents to in-hospital cardiopulmonary arrest. Chest, 98, 13881392.
  • Smith, G. (2011, 7 February). Improving the detection and response to patient deterioration. Retrieved October 27, 2012, from http://www.nice.org.uk/usingguidance/sharedlearningimplementingniceguidance/examplesofimplementation/eximpresults.jsp?o = 427
  • Smith, G. B., Prytherch, D. R., Watson, D., Forde, V., Windsor, A., Schmidt, P. E., & Meredith, P. (2012). S(p)O(2) values in acute medical admissions breathing air—Implications for the British Thoracic Society guideline for emergency oxygen use in adult patients? Resuscitation, 83, 12011205.
  • Vermeulen, H., Storm-Versloot, M. N., Goossens, A., Speelman, P., & Legemate, D. A. (2005). Diagnostic accuracy of routine postoperative body temperature measurements. Clinical Infectious Diseases, 40, 14041410.
  • VMS Veiligheids programma. (2008). Vroege herkenning en behandeling van de vitaal bedreigde patiënt. [Early detection and treatment of critically ill patients]. Retrieved from http://www.vmszorg.nl/themas/vitaal-bedreigde-patient
  • VMS Veiligheids programma. (2009). Voorkomen van lijnsepsis en behandeling van ernstige sepsis [Prevention of line sepsis and treatment of severe sepsis]. Retrieved from http://www.vmszorg.nl/themas/sepsis
  • Whiting, P., Rutjes, A. W., Reitsma, J. B., Bossuyt, P. M., & Kleijnen, J. (2003). The development of QUADAS: A tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BioMed Central Medical Research Methodology. doi:10.1186/1471-2288-3-25
  • Zeitz, K., & McCutcheon, H. (2006). Observations and vital signs: Ritual or vital for the monitoring of postoperative patients? Applied Nursing Research, 19, 204211.

Supporting Information

  1. Top of page
  2. Abstract
  3. Purpose
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Disclaimer: Supplementary materials have been peer-reviewed but not copyedited.

FilenameFormatSizeDescription
jnu12048-sup-0001-TableS1.doc51KTable S1. Quality assessment
jnu12048-sup-0002-Figures.tif1676KFigure 1. Flowchart
jnu12048-sup-0003-SuupMat.doc47KAddendum 1. Search strategies by database
jnu12048-sup-0004-SuppMat.doc45KAddendum 2. Selection criteria
jnu12048-sup-0005-SuppMat.doc40KAddendum 3. Quality assessment tools

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.