To evaluate the clinical validity of an instrument (called the Triage Score) for rating the severity of patients needs for respiratory care, we compared Triage Score ratings with clinical features and outcomes in 2 independent groups of adult, non-ICU inpatients (Group 1 = 98 and Group 2 = 179) receiving respiratory care at the Cleveland Clinic Hospital between January 1992 and April 1994. The Score was originally developed to help identify patients in whom respiratory therapy could be safely deferred when orders for respiratory care outstripped the availability of therapists to deliver care. Ratings are based on a 32-point ordinal score in which the patient is rated on 8 axes relevant to respiratory care needs (history of pulmonary disease, surgical status on current admission, chest radiograph, respiratory pattern, mental status, cough, and level of activity); and the scores are collapsed into 5 Triage Score categories, with Triage 1 indicating the highest severity of respiratory illness. For both patient groups evaluated, Triage Score ratings showed consistent and significant relationships (p < 0.05) with patient age, FIO2 requirement, length of hospital stay, days requiring respiratory therapy, and number of respiratory therapy modalities used. Also, respiratory therapy charges correlated significantly (p < 0.001) with Triage Score ratings in both groups, and the mortality rate within the larger patient group (Group 2) tended to increase with lower Scores (p = 0.07). Although the instrument is not intended for mortality prediction, Receiver Operating Characteristic analysis of Triage Score ratings for predicting mortality showed an area under the curve of 0.637 (p = 0.058), suggesting that ratings tracked indicators of worsened clinical status. Finally, analysis of Triage Score assignments by 2 separate raters indicated moderate agreement beyond chance (observed agreement in 62% and 90% of patients in the 2 groups). Based on significant and reproducible relationships between Triage Score ratings and relevant clinical measures in 2 independent patient groups, we believe that the Triage Score is valid and reliable in evaluating patients’ respiratory care needs in a therapist-driven protocol program.
Get full access to this article
View all access options for this article.
References
1.
KnausWA, ZimmermanJE, WagnerDP, DraperEA, LawrenceDE. APACHE-acute physiology and chronic health evaluation: A physi-ologically based classification instrument. Crit Care Med1981; 9(8):591–597.
2.
TeresD, LemeshowS, AvruninJS, PastidesH.Validation of the mor-tality prediction model for ICU patients. Crit Care Med1987; 15(3):208–213.
3.
ZibrakJD, RossettiP, WoodE.Effect of reductions in respiratory therapy on patient outcome. N Engl J Med1986; 315(5):292–295.
4.
TorringtonKG, HendersonCJ. Perioperative respiratory therapy (PORT). A program of preoperative risk assessment and individu-alized postoperative care. Chest1988; 93(5):946–951.
5.
StollerJK, SkibinskiCI, GilesDK, KesterL.Physician-ordered res-piratory care vs. physician-ordered use of a respiratory therapy con-sult service. Results of a prospective observational study. Chest1996; 110(2):422–429.
6.
GilesDK, KesterL, StollerJK. Validation of the triage rating instru-ment in the respiratory therapy consult service (abstract). Respir Care1994; 39:1096.
StollerJK, HaneyD, BurkhartJ, FergusL, GilesD, HoisingtonE, KesterL, KomaraJ, McCarthyK, MeredithR, OrensD, and the Section of Respiratory Therapy. Physician-ordered res-piratory care vs. physician-ordered use of a respiratory therapy con-sult service: early experience at the Cleveland Clinic Foundation. Respir Care1993; 38:1143–1154.
9.
KramerMS, FeinsteinAR. Clinical biostatistics: LIV. The biostatistics of concordance [Published erratum Clin Pharmacol Ther 1989; 46(3):309]. Clin Pharmacol Ther1981; 29(1):111–123.
10.
CicchettiDV, AllisonT.A new procedure for assessing reliabil-ity of scoring EEG sleep recordings. Am J EEG Technology1971; 11:101–109.
11.
DorfmanDD, AlfE Jr. Maximum likelihood estimation of param-eters of signal detection of confidence intervals-rating method data. J Math Psych1969;6:487–496.
