Article, Emergency Medicine

Soft tissue oxygen saturation to predict admission from the emergency department: A prospective observational study

a b s t r a c t

Objective: We evaluated a soft tissue oxygen saturation (Sto2) measurement at triage for predicting admission to the hospital in adults presenting to the emergency department (ED) in addition to data routinely gathered at tri- age.

Methods: This was a prospective, observational, single center study of adults presenting to the ED for evaluation. Research assistants obtained thenar eminence Sto2 measurements on subjects in ED triage. ED providers not in- volved in the study then made all management and Disposition decisions. We prospectively collected data on each subject’s final ED disposition (admission versus discharge). We identified the optimal Sto2 cutoff value for predicting admission. We then used logistic regression modeling to describe the added predictive value of Sto2 beyond routinely collected triage data including Emergency Severity Index level, age, and vital signs.

Results: We analyzed 2588 adult (N 17 years) subjects with 743 subjects (28.7%) admitted to the hospital. Sto2 b 76% was the optimal diagnostic cutoff for predicting admission. Of subjects with Sto2 b 76%, 158 of 384 (41.1%) underwent admission versus 585 of 2204 (26.5%) subjects with Sto2 >= 76. After controlling for age, vital signs, and ESI level in the logistic regression analysis, Sto2 b 76% had an odds ratio of 1.54 (95% confidence interval (CI), 1.19 to 1.98) for predicting admission.

Conclusions: Sto2 may provide additional prognostic data to routine Triage assessment regarding the disposition for undifferentiated adult patients presenting to the ED.

  1. Introduction
    1. Background

Near-infrared spectroscopy (NIRS) uses light spectroscopy to evalu- ate relative amounts of oxyhemoglobin and deoxyhemoglobin. This technology has mostly been studied during resuscitation to identify re- gional tissue hypoxia as a potential early precursor to shock before sys- temic signs or lab abnormalities are present [1]. Decreased Sto2 measurements in various trauma populations have been associated with a range of clinical outcomes to include development of multi- organ dysfunction syndrome (MODS), requirement for blood product transfusion, or requirement of emergent surgery [2-6]. Sto2 b 75% has

Abbreviations: Sto2, soft tissue oxygen saturation; NIRS, near-infrared spectroscopy; MODS, multi-organ dysfunction syndrome; ESI, Emergency Severity Index; CI, confidence interval.

* Corresponding author at: San Antonio Military Medical Center, Department of

Emergency Medicine, 3841 Roger Brooke Dr, Fort Sam Houston, TX 78234, United States.

E-mail address: [email protected] (W.T. Davis).

been associated with the development of MODS in ED trauma patients [7]. A meta-analysis of Sto2 measurements at the thenar eminence found that patients with sepsis had lower Sto2 values [8]. Sto2 b 75% after resuscitation is associated with increased mortality in patients with severe sepsis or septic shock [9]. Limited data is available regarding this technology’s utility for prognosticating outcomes among undiffer- entiated ED patients who have not yet undergone resuscitation.

Importance

Accurate ED triage is essential to guiding timely care and Resource management in the setting of high patient demand for ED resources [10]. Sto2 is a rapid, noninvasive test which could provide prognostic in- formation in addition to that provided by validated ED triage assess- ment tools such as the Emergency Severity Index [11,12].

Goals of this investigation

The objective of this study was to evaluate the additional prognostic information provided by a single Sto2 measurement at triage in the

http://dx.doi.org/10.1016/j.ajem.2017.03.013 0735-6757/

setting of routine triage data to identify adult patients presenting to the ED requiring hospital admission. In accordance with the existing Sto2 literature, we anticipated that utilization of Sto2 as a binary variable would provide better prognostic value than as a continuous variable [7,9,13]. Therefore, our a priori goal was to determine the optimal Sto2 cut-off value for identifying subjects requiring admission and use this value to analyze the Predictive power of Sto2 beyond that provided by routine triage data. We hypothesized that patients with Sto2 values below the optimal cut-off would be more likely to undergo hospital ad- mission than patients with Sto2 values above the optimal cut-off even after controlling for other routinely collected triage data such as vital signs, age, and ESI.

  1. Methods
    1. Study design and setting

This was a prospective, observational, single center study conducted at San Antonio Uniformed Services Health Education Consortium. This facility is an urban tertiary care hospital serving active duty military per- sonnel, retirees, and beneficiaries in the San Antonio metropolitan area. The annual ED census during the study period was 76.959 patients. The hospital institutional review board approved the study protocol. We ad- hered to strengthening the Reporting of Observational Studies in Epide- miology (STROBE) Statement guidelines in our research design, reporting, and analysis [14].

Selection of participants

We enrolled a convenience sample of adult patients presenting to the ED from November 2012 to November 2013. All adults (N 17 years old) presenting to the ED during this time period were potentially eligi- ble for inclusion. We excluded patients with cardiac arrest, major trau- ma activation, peripheral vascular disease, amputated upper extremities, skin abnormalities at the thenar eminence, or clinical evi- dence of intoxication.

Two research assistants screened adults presenting to the ED and obtained consent from all subjects during 112 shifts covering 8 h pe- riods when research assistant staffing was available from Monday through Friday between the hours of 8 AM and midnight. Research as- sistants collected an Sto2 reading at triage and recorded the chief com- plaint and initial vital signs for each subject. Triage nurses assigned each subject an ESI score as part of routine ED triage processes using ESI ver- sion 4, (Agency for Healthcare Research and Quality, Rockville, Mary- land) [15]. In the event that the patients actively undergoing triage outnumbered the research assistants, we instructed the research assis- tants to screen the most recently triaged patient. Research assistants prospectively collected data on each subject to include discharge diag- nosis, and subject disposition (binary variable: admission versus dis- charge). All enrolled subjects received routine care from our institution. Treating providers were not privy to the Sto2 result, so these data were not available for diagnostic decision making.

Methods of measurement

Research assistants applied the tissue saturation oximeter, (Inspectra Model 300 Sto2 Spot Check, Hutchinson Technology Inc., Hutchinson MN) to the thenar eminence of each participant. The re- search assistants obtained these measurements in a standardized fash- ion. Specifically, patients were in a relaxed position several minutes prior to this measurement and held the measured arm flat on the bed during the Sto2 measurement to reduce variability in measurements. This handheld device uses NIRS technology to measure an approximate percent oxygen saturation (0-99%) of the hemoglobin circulating in the thenar muscle [16,17]. We chose the thenar eminence given its previous validation as a Sto2 measurement site [7,18,19].

We coded free text chief complaint data recorded by nursing staff in accordance with the most common reasons for ED visits from the Na- tional Hospital Ambulatory Care Survey (NHACS) [20]. We coded final diagnosis data according to primary diagnoses from the NHACS [20].A second evaluator used a random number generator and coded 10% of the data. We calculated measures of inter-rater reliability (kappa coeffi- cients) for both of these variables.

Outcome measures

Our primary outcome was the binary variable of subject disposition from the ED: admission versus discharge. We defined admission as sub- ject transport from the ED to an inpatient care floor or to another hospi- tal for further care. We stored all data in a secure Excel database (version 14; Microsoft, Redmond, WA).

Data analysis

The primary analysis evaluated the test characteristics of Sto2 to pre- dict admission to the hospital. We described continuous variables with means, ordinal variables with medians, and nominal variables with fre- quencies or percentages. We used bootstrapping techniques to calculate 95% confidence intervals (CIs) around main outcomes and differences in baseline characteristics [21]. We performed multivariate logistic regres- sion on the decision of patient admission. We added terms in a step- wise fashion using forward model-selection procedure to keep the model as parsimonious as possible and to assess the stability of the sta- tistical estimates across multiple model specifications [22]. We included Sto2, age, temperature, heart rate, blood pressure, and ESI as explanato- ry terms. We treated Sto2 as a curved exponential parameter and calcu- lated the profile log-likelihood for all cutoff values from 0% to 99%. We constructed a plot of the resulting curve to evaluate the appropriateness of analyzing Sto2 as a continuous variable (indicated by a plateau) ver- sus a binary variable with evidence of an optimal cut-off value (indicat- ed by a peak) [23]. We used R (version 3.3.1, Foundation for Statistical Computing, Vienna, Austria) for all data analysis [24]. We excluded sub- jects with missing data from the final analysis.

Our sample size estimate assumed ? = 0.05, ? = 0.20, and two-

sided statistical testing. We powered our study to detect a 5% difference in probability of admission for subjects with Sto2 values above versus below the diagnostic cutoff. We anticipated a 17% admission rate based on internal performance improvement data. Our estimated sam- ple size was 2046 subjects.

  1. Results
    1. Characteristics of study subjects

Approximately 6430 patients presented to the ED during the 112 shifts of active enrollment. We calculated this value by subtracting total pediatric visits from our total ED census during the study period, (76,595 minus 13,671), and multiplying this result by hours of active enrollment divided by total hours of the study period. Ultimately, 2968 screened subjects were eligible for the study. Of the eligible sub- jects, 326 declined consent, and we excluded 54 subjects due to missing data. We included 2588 subjects in the final analysis (Fig. 1). Seven hun- dred forty-three subjects (28.7%) underwent admission. Older subjects and males were more likely to experience admission (Table 1). Mean Sto2 was 79.4% in admitted patients (standard deviation 5.3%) com- pared to 80.4% in not admitted patients (standard deviation 4.8%). When stratified by ESI level, the median Sto2 values were lower in ad- mitted patients than not admitted patients across each ESI level with more divergence noted in patients with ESI 1 or 5 (Fig. 2).

with Sto2 b 76%. After controlling for age, vital signs, and ESI level in the logistic regression analysis, Sto2 b 76% had an odds ratio of 1.54 (95% CI, 1.19 to 1.98) for predicting admission.

Sto2 had the following characteristics as a solitary test from ED tri- age for predicting admission to the hospital. Sto2 b 76% had a sensitivity of 87.8% (95% CI, 86% to 89%), specificity of 21.3% (95% CI, 18% to 24%),

positive likelihood ratio of 1.12 (95% CI, 1.07 to 1.16), negative likeli-

hood ratio 0.57 (95% CI, 0.48 to 0.68), positive predictive value of

50.0% (95% CI, 36% to 46%) and negative predictive value of 73.5% (95% CI, 72% to 75%) for identifying subjects ultimately admitted.

Kappa coefficients were 0.92 and 0.98 for chief complaint and final diagnosis respectively between the two coders for chief complaint and final diagnosis data. Sto2 b 76% was a significant predictor of admission in chief complaints of chest pain, shortness of breath, fever, and abdom- inal pain (Appendix Table 3). Similarly, Sto2 b 76% was a significant pre- dictor of admission in subjects with final diagnoses of diseases of the circulatory system, diseases of the respiratory system, and ill-defined signs and symptoms (Appendix Table 4).

  1. Discussion

3.2. Main results

Fig. 1. subject enrollment and exclusion.

This study evaluated the use of a single noninvasive soft tissue oxim- etry measurement as a triage tool to identify adults presenting to the ED likely to undergo admission to the hospital. In this large, prospective, observational, single center study, the use of Sto2 demonstrated poten- tial utility for clinical prognostication in addition to routine measure- ments at triage. This study adds to previous literature as a large data set examining Sto2 in an undifferentiated ED adult population that is not actively undergoing resuscitation.

Optimal Diagnostic thresholds identified for Sto2 measurements at the thenar eminence range from 65% to 75% in the literature depending on the patient population and outcome of interest [6,7,25]. Sto2 b 70% at

We identified a diagnostic cutoff value of Sto2 b 76% as indicated by the peak at Sto2 value b 76% (Fig. 3), meaning that the Sto2 values best explain the decision to admit when using a cutoff value of b 76%.

One hundred and fifty-eight of 384 (41.1%) subjects with Sto2 b 76% experienced admission compared to 585 of 2204 (26.5%) subjects with Sto2 >= 76%, yielding a relative risk of 1.5 (95% CI, 1.3 to 1.8). Subjects with Sto2 b 76% were older and more likely to be female (Appendix Table 1). Relative risk of admission for Sto2 b 76% stratified by ESI level is available in Appendix Table 2 and graphically in Appendix Fig. 1. Table 2 displays the output of the logistic regression model for the binary variable corresponding to the decision to admit with admission corresponding to a true value. A model with Sto2 as the only explanato- ry variable yields a positive and statistically significant estimate of 0.66, which corresponds with an odds ratio of 1.93 for admission in patients

Table 1

Baseline characteristics of admitted and not admitted subjects.

ED triage in a study of cancer patients presenting with SIRS criteria or hypotension had an odds ratio of 2.6 (95% CI, 1.2 to 5.9) for admission to the ICU [13]. A sepsis screening tool at ED triage composed of Sto2 b 75% and initial vital signs had a sensitivity of 85.7% and specificity of 78.4% for identifying sepsis in non-trauma adult patients presenting to the ED [26]. In a study of septic patients at ED triage, Sto2 b 70% lacked prognostic value in identifying patients who would later develop severe sepsis [27].

Our study found that an Sto2 cut-off of b 76% had the greatest dis- criminatory power for predicting hospital admission in our large dataset of general ED patients. This cutoff value is consistent with prior litera- ture analyzing Sto2 b 75% as a prognostic measurement for mortality in septic patients after resuscitation and for development of MODS in trauma patients [7,9]. Subjects with Sto2 b 76% had an odds ratio of

1.54 for admission after controlling for age, ESI, and vital signs in the lo- gistic regression analysis. This model takes into account data which is routinely gathered at triage, and Sto2 b 76% still provided additional prognostic value in identifying patients who would require admission.

Variablea Admitted subjects (N = 743)

Not admitted subjects (N = 1845)

This finding indicates that Sto2 provides information beyond simply identifying patients that already appear sick on arrival by current triage

Age, y 64 (50, 78) 47 (30, 63)

Sex, male (%) 49 44

SBP, mm Hg 138 (122, 155) 134 (121, 138)

DBP, mm Hg 79 (69, 90) 80 (71, 88)

HR 82 (71, 97) 83 (73, 95)

RR 18 (18, 20) 18 (18, 20)

Temp 98 (98, 98) 98 (98, 98)

protocols.

Our study analyzed an undifferentiated ED patient population. When data was stratified by chief complaint, Sto2 had statistically sig- nificant predictive power in abdominal pain, chest pain, fever and short- ness of breath. Given that ED triage is a balance of gathering data and time efficiency, Sto2 may prove to be most useful at triage only for spe-

Pulse

oximetry Soft tissue

O2%

98 (96, 99) 98 (97, 100)

80 (76, 83) 80 (78, 84)

cific chief complaints. Our study shows that a single Sto2 measurement obtained at triage can provide additional prognostic data in excess of data obtained via routine Triage protocols. Future studies of Sto2 within

targeted subsets of our study population, such as patients presenting

Abbreviations: CI = confidence interval y = year, SBP = systolic blood pressure, DBP = diastolic blood pressure, HR = heart rate, RR = respiratory rate.

a All data are presented as median values followed by the interquartile range in pa- rentheses where applicable.

with chest pain, fever or shortness of breath, may improve the diagnos- tic characteristics of Sto2 for identifying patients who will require ad- mission to the hospital.

Fig. 2. Probability density functions of soft tissue oximetry values in admitted and non-admitted subjects stratified by Emergency Severity Index levels.

Limitations

Our study has several limitations. Our primary outcome of admission to the hospital reflects the decision of an individual healthcare provider rather than a physiologic outcome. To the extent that observational data supports higher mortality rates in admitted patients compared to non- admitted patients, the decision to admit may serve as a surrogate for more definitive outcome measures [28,29]. Nevertheless, we lack follow up data on the study population, so we are unable to establish the utility of Sto2 in predicting outcomes such as mortality.

Our study may have limited generalizability as it was a convenience sample based on research assistant availability during weekdays in a single center study. Selection bias may exist as approximately 46% of

Fig. 3. Profile log likelihood of best fit for different Sto2 cut-off values within the logistic regression model with higher values corresponding to a better fit.

patients presenting to the ED during active enrollment were screened for the study. We sought to minimize this selection bias by instructing research assistants to follow a systematic approach of screening the most recently triaged patients when unable to enroll every eligible pa- tient. We lack data on the characteristics of all patients presenting to our ED during the study period not recruited into the study. Conse- quently, we cannot establish whether our study population is represen- tative of the broader population of all ED patients.

We only studied a single NIRS device so our results may not apply to the use of other NIRS devices [19]. It is unclear whether replication of our study in alternative settings would identify a similar association be- tween low Sto2 and the ED provider decision to admit patients to the hospital. Moreover, it is not certain that the use of a cut-off value neces- sarily always provides optimal prognostic information. Our profile like- lihood analysis suggested that in our dataset analysis of StO2 as a binary variable was superior to analysis as a continuous variable assuming a

Table 2 Logistic regression output for variables as predictors of hospital admission from the emer- gency department.

Variable

Natural log of odds ratio

Standard error

Odds ratio (95% CI)

(Intercept)

-0.80

0.94

0.45 (0.07, 2.84)

Sto2 b 76%

0.43

0.13

1.54 (1.19, 1.98)

Age N 45

0.61

0.16

1.84 (1.35, 2.52)

Age N 55

0.42

0.16

1.52 (1.11, 2.08)

Age N 65

0.45

0.13

1.57 (1.22, 2.02)

Temperature N 102

1.40

0.91

4.06 (0.68, 24.13)

Heart rate N 100

0.35

0.13

1.42 (1.10, 1.83)

Diastolic BP N 100

0.31

0.18

1.36 (0.96, 1.94)

ESI 2

-0.04

0.94

0.96 (0.15, 6.06)

ESI 3

-1.17

0.94

0.31 (0.05, 1.96)

ESI 4

-2.87

0.97

0.06 (0.01, 0.38)

ESI 5

-1.59

1.05

0.20 (0.03, 1.60)

Base cases for dummy terms are as follows: StO2 >= 76, Age b 46, Temperature <= 102, Heartrate b 101, Diastolic BP b 101, and ESI = 1.

linear relationship with the log odds of admission. However, these find- ings may not hold in other settings. Moreover, it is possible that alterna- tive advanced statistical techniques may demonstrate predictive utility in analyzing Sto2 as a continuous variable without the assumption of a linear relationship with log odds of admission.

Our study was observational and so is susceptible to confounding. Our logistic regression model suggests that of the variables we collected, Sto2 measurement added prognostic information regarding whether patients would undergo hospital admission when controlling for other measured study variables. Nevertheless, we are unable to exclude un- measured confounders.

  1. Conclusions

Sto2 is a non-invasive measurement which added additional prog- nostic information regarding need for admission compared to routine triage data in our undifferentiated ED patient population at triage. Fu- ture Research priorities should be to replicate our study in other ED en- vironments assessing Sto2 within specific chief complaints such as chest pain or shortness of breath while measuring more terminal outcome measures such as hospitalization days and mortality. Our prospective, observational study establishes a diagnostic threshold for Sto2 predicting admission and establishes Sto2 measurements as a safe and simple tool that may provide additional prognostic information to conventional triage methods in undifferentiated adults presenting to the ED.

Conflict of interest disclosure

WTD, JL, RMB, JH, SGS, TBS, MDA report no conflicts of interest.

Prior presentations

ACEP Research Forum, October 28, 2014, Chicago, IL, USA.

Funding sources

This work was supported by the Army Medical Department Ad- vanced Medical technology Initiative [grant award number 2101].

Disclaimers

The view(s) expressed herein are those of the author(s) and do not reflect the official policy or position of Brooke Army Medical Center, the

U.S. Army Medical Department, the U.S. Army Office of the Surgeon General, the Department of the Army, the Department of the Air Force and Department of Defense or the U.S. Government. This original contri- bution has not been published, it is not under consideration for publica- tion elsewhere, its publication is approved by all authors and tacitly or explicitly by the responsible authorities where the work was carried out, and that, if accepted, it will not be published elsewhere in the same form, in English or in any other language, including electronically without the written consent of the copyright-holder.

Appendix A

Appendix Table 1

Baseline characteristics of subjects stratified by soft tissue oximetry values.

Variablea

Sto2 b 76% (N = 384)

Sto2 >= 76% (N = 2204)

Median age, y

64 (42, 81)

52 (33, 66)

Sex, male (%)

39

47

SBP, mm Hg

135 (120, 151)

133 (122, 151)

DBP, mm Hg

78 (69, 87)

80 (71, 89)

HR

81 (71, 95)

83 (72, 96)

RR

18 (18, 20)

18 (18, 20)

Temp

98 (98, 98)

98 (98, 98)

Pulse oximetry

98 (96, 99)

98 (97, 99)

a All data are presented as median values followed by the interquartile range in pa- rentheses where applicable.

Appendix Table 2

Relative risk of admission for Sto2 b 76% stratified by Emergency Severity Index level.

ESI level

Sto2 b 76%

Sto2 >= 76%

Relative risk of admission for ST O2 b 76% (95% CI)

N admitted

N total

N admitted

N total

ESI 1

1

1

3

5

1.7 (+1 to infinity)

ESI 2

61

90

200

380

1.3 (+1.1 to 1.5)

ESI 3

90

247

365

1539

1.5 (+1.3 to 1.9)

ESI 4

2

37

15

255

0.8 (+0 to +2.9)

ESI 5

4

9

2

25

5.5 (+1 to infinity)

All patients

158

384

585

2204

1.5 (+1.3 to +1.8)

Appendix Table 3

Relative risk of admission for Sto2 b 76% stratified by presenting chief complaint.

Principle reason for visit

Sto2 b 76%

Sto2 >= 76%

Relative risk of admission for StO2 <= 75% (95% CI)

N admitted (N = 158)

N total

(N = 384)

N admitted (N = 585)

N total

(N = 2204)

Stomach and abdominal pain, cramps and spasms

15

38

82

334

1.6 (1.0 to 2.1)

Chest pain and related symptoms

32

48

129

310

1.5 (1.2 to 1.8)

Fever

6

8

13

43

2.0 (1.2 to 3.3)

Headache

3

8

16

105

2.2 (0 to 4.68)

Back symptoms

0

4

10

75

0.0 (0 to 0)

(continued on next page)

Appendix Table 3 (continued)

Principle reason for visit

Sto2 b 76%

Sto2 >= 76%

Relative risk of admission for StO2 <= 75% (95% CI)

N admitted (N = 158)

N total

(N = 384)

N admitted (N = 585)

N total

(N = 2204)

Shortness of breath

27

41

56

129

1.4 (1.08 to 1.66)

Cough

4

7

19

72

1.96 (0.6 to 3.5)

Vomiting

0

6

6

29

0 (0 to 0)

Pain, site not referable to a specific body system

3

11

9

62

1.7 (0 to 3.4)

Symptoms referable to throat

2

8

8

61

1.7 (0 to 4.0)

Nausea

3

15

9

69

1.4 (0 to 2.9)

Accident, not otherwise specified

9

17

11

40

1.5 (0.88 to 2.2)

Motor vehicle accident, type of injury unspecified

0

2

4

38

0.0 (0 to 0)

Vertigo – dizziness

1

10

19

87

0.49 (0 to 1.5)

Leg symptoms

5

17

18

126

1.8 (0.5 to 3.2)

skin rash

1

12

15

68

0.42 (0 to 1.4)

Injury, other and unspecified type

2

10

9

38

0.87 (0 to 2.0)

Low back symptoms

1

9

9

75

0.93 (0 to 3.2)

All other reasons

44

113

143

443

1.1 (0.91 to 1.4)

Appendix Table 4

Relative risk of admission for Sto2 b 76% stratified by primary final diagnosis from the emergency department.

Major disease category

Sto2 b 76%

Sto2 >= 76%

Relative risk of admission for StO2 <= 75% (95% CI)

N admitted (N = 158)

N total

(N = 384)

N admitted (N = 585)

N total

(N = 2204)

Infectious and parasitic diseases

8

12

9

28

1.57 (0.9 to 2.3)

Endocrine, nutritional, metabolic diseases and immunity disorders

5

13

23

77

1.2 (0.4 to 2.1)

Mental disorders

2

6

12

37

1.0 (0 to 2.4)

Disease of the nervous system and sense organs

4

14

31

127

1.2 (0.3 to 2.2)

Diseases of the circulatory system

31

43

87

167

1.3 (1.1 to 1.5)

Diseases of the respiratory system

16

31

56

207

1.7 (1.1 to 2.3)

Diseases of digestive system

16

41

77

242

1.2 (0.7 to 1.6)

Diseases of the Genitourinary system

18

51

59

269

1.5 (1.0 to 2.0)

Diseases of the skin and subcutaneous tissue

2

19

16

85

0.6 (0 to 1.6)

Diseases of the musculoskeletal system and connective tissue

3

18

8

126

2.0 (0 to 4.3)

Symptoms, signs, and ill-defined conditions

45

107

179

689

1.5 (1.2 to 1.9)

Injury and Poisoning

7

20

22

126

1.8 (0.8 to 2.8)

All other diagnoses

1

9

6

24

0.5 (0 to 1.7)

Appendix Fig. 1. Comparison of patient admissions with a Sto2 b 76% cut-off value stratified by Emergency Severity Index levels.

References

  1. Ward KR, Ivatury RR, Barbee RW, Terner J, Pittman R, Filho IP, et al. Near infrared spectroscopy for evaluation of the trauma patient: a technology review. Resuscita- tion 2006;68(1):27-44.
  2. Duret J, Pottecher J, Bouzat P, Brun J, Harrois A, Payen JF, et al. Skeletal muscle oxy- genation in Severe trauma patients during haemorrhagic shock resuscitation. Crit Care 2015;19:141.
  3. Beekley AC, Martin MJ, Nelson T, Grathwohl KW, Griffith M, Beilman G, et al. Contin- uous noninvasive tissue oximetry in the early evaluation of the combat casualty: a prospective study. J Trauma 2010;69(Suppl. 1):S14-25.
  4. Nicks BA, Campons KM, Bozeman WP. Association of low non-invasive near-in- frared spectroscopic measurements during initial trauma resuscitation with fu- ture development of Multiple organ dysfunction. World J Emerg Med 2015; 6(2):105-10.
  5. Carlile C, Wade CE, Baraniuk MS, Holcomb JB, Moore LJ. Evaluation of sto2 tissue per- fusion monitoring as a tool to predict the need for lifesaving interventions in trauma patients. Am J Surg 2015;210(6):1070-5 (discussion 5).
  6. Khasawneh MA, Zielinski MD, Jenkins DH, Zietlow SP, Schiller HJ, Rivera M. Low tis- sue oxygen saturation is associated with requirements for transfusion in the rural trauma population. World J Surg 2014;38(8):1892-7.
  7. Cohn SM, Nathens AB, Moore FA, Rhee P, Puyana JC, Moore EE, et al. Tissue oxygen saturation predicts the development of organ dysfunction during traumatic shock resuscitation. J Trauma 2007;62(1):44-54 (discussion-5).
  8. Neto AS, Pereira VG, Manetta JA, Esposito DC, Schultz MJ. Association between static and dynamic thenar near-infrared spectroscopy and mortality in patients with sep- sis: a systematic review and meta-analysis. J Trauma Acute Care Surg 2014;76(1): 226-33.
  9. Vorwerk C, Coats TJ. The prognostic value of tissue oxygen saturation in emer- gency department patients with severe sepsis or septic shock. Emerg Med J 2012;29(9):699-703.
  10. Fernandes CM, Tanabe P, Gilboy N, Johnson LA, McNair RS, Rosenau AM, et al. Five- level triage: a report from the acep/ena five-level triage task force. J Emerg Nurs 2005;31(1):39-50 (quiz 118).
  11. Wuerz RC, Milne LW, Eitel DR, Travers D, Gilboy N. Reliability and validity of a new

five-level triage instrument. Acad Emerg Med 2000;7(3):236-42.

  1. Tanabe P, Gimbel R, Yarnold PR, Kyriacou DN, Adams JG. Reliability and validity of scores on the emergency severity index version 3. Acad Emerg Med 2004;11(1): 59-65.
  2. Bazerbashi H, Merriman KW, Toale KM, Chaftari P, Cruz Carreras MT, Henderson JD, et al. Low tissue oxygen saturation at emergency center triage is predictive of inten- sive care unit admission. J Crit Care 2014;29(5):775-9.
  3. von Elm EAD, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. The strengthening the reporting of observational studies in epidemiology (strobe) state- ment: guidelines for reporting observational studies. PLoS Med 2007;4(10):e296.
  4. Gilboy Tanabe, Travers Rosenau. Emergency Severity Index : a triage tool for emergency department care, version 4. Implementation handbook 2012 edition. Rockville, MD: Agency for Healthcare Research and Quality; 2011 Ahrq publication no. 12-0014.
  5. Hutchinson Technology Inc. Inspectra sto2 spot check, model 300; 2012 1-4.
  6. Hutchinson Technology Inc. Inspectra sto2 spot check system, 510(k) summary, k103613; 2011.
  7. Crookes BA, Cohn SM, Bloch S, Amortegui J, Manning R, Li P, et al. Can near-infrared spectroscopy identify the severity of shock in trauma patients? J Trauma 2005; 58(4):806-13 (discussion 13-6).
  8. Scheeren TW, Schober P, Schwarte LA. Monitoring Tissue oxygenation by near infra- red spectroscopy (nirs): background and current applications. J Clin Monit Comput 2012;26(4):279-87.
  9. Pitts SR, Niska RW, Xu J, Burt CW. National hospital ambulatory medical care survey: 2006 emergency department summary. Natl Health Stat Rep 2008;7:1-38.
  10. Efron B. Better bootstrap confidence intervals. J Am Stat Assoc 1987;82(397): 171-85.
  11. Hadley W. The split-apply-combine strategy for data analysis. J Stat Softw 2011; 40(1):1-29.
  12. Davison A, Padoan S, Ribatet M. statistical modeling of spatial extremes. Stat Sci 2012:161-86.
  13. R Core Team. R: a language and environment for statistical computing. Vienna, Aus- tria: R Foundation for Statistical Computing; 2016.
  14. Iyegha UP, Conway T, Pokorney K, Mulier KE, Nelson TR, Beilman GJ. Low sto2 mea- surements in surgical intensive care unit patients is associated with poor outcomes. J Trauma Acute Care Surg 2014;76(3):809-16.
  15. Goerlich CE, Wade CE, McCarthy JJ, Holcomb JB, Moore LJ. Validation of sepsis screening tool using sto2 in emergency department patients. J Surg Res 2014; 190(1):270-5.
  16. Goulet H, Andre S, Sahakian GD, Freund Y, Khelifi G, Claessens YE, et al. Accuracy of oxygen tissue saturation values in assessing severity in patients with sepsis admit- ted to emergency departments. Eur J Emerg Med 2014;21(4):266-71.
  17. Dean NC, Jones JP, Aronsky D, Brown S, Vines CG, Jones BE, et al. Hospital admission decision for patients with community-acquired pneumonia: variability among phy- sicians in an emergency department. Ann Emerg Med 2012;59(1):35-41.
  18. Labarere J, Stone RA, Obrosky DS, Yealy DM, Meehan TP, Fine JM, et al. Comparison of outcomes for low-risk outpatients and inpatients with pneumonia: a propensity-ad- justed analysis. Chest 2007;131(2):480-8.

Leave a Reply

Your email address will not be published. Required fields are marked *