Evaluation of An After-Hours Call Center: Are Pediatric Patients Appropriately Referred to the Emergency Department?

Friday, October 25, 2013: 4:45 PM
Windermere Ballroom Y (Hyatt Regency Orlando, formerly the Peabody)
Kaynan Doctor, MD, Children's National Medical Center, Washington, DC

Purpose

Pediatric after-hours nurse telephone triage (AHNTT) systems are nurse-lead services that utilize computerized algorithms to advise care and disposition of pediatric patients between normal clinic hours. AHNTTs have become a widespread convenience employed by numerous practices and have helped standardize care and improve the lifestyle of outpatient pediatric physicians. There is concern, however, that AHNTT protocols may be responsible for over-referring non-emergent patients to the emergency department (ED), thus inundating an already over-burdened service. Our institution employs a non-pediatric hospital based call center which is similar to many outpatient offices nationwide. The objectives of this study were to critically review this AHNTT and identify those algorithms responsible for non-essential referrals.

Methods

This retrospective observational study was performed at a tertiary medical care facility over one year. AHNTT forms of patients aged under 18 years who were exclusively referred by triage nurses without any influence from any physicians (using the Barton Schmitt protocols) were reviewed and their ED course evaluated by consulting the electronic medical record. ‘Essential’ referrals to the ED were classified by a panel of physicians as presentations warranting either immediate evaluation, or referrals requiring ‘essential interventions’ such as serum lab tests, imaging, complex procedures, IV medications, consultation or admission.

Results

220 patients were included in this study.  Of these, 73 (33%) were classified non-essential while 147 (67%) were classified as essential. Non-essential referrals were significantly younger compared to essential referrals (p <0.05 ). Non-essential referrals also had lower triage scores (p=0.026) and shorter ED stays (p<0.0001). The AHNTT algorithms for ‘Fever - 3 months or Older’ (12.3%), ‘Vomiting Without Diarrhea (8.2%) ‘Trauma - Head’ (8.2%), ‘Headache’ (6.8%) and ‘Sore Throat’ (5.5%) were determined most likely to result in a non-essential referral.

Conclusion

Our study identifies a significant proportion of unnecessary pediatric visits to the ED as a result of this AHNTT service. A younger population in the non–essential group may reflect lower comfort levels in managing sick infants or toddlers over the phone by non-pediatric personnel. We recommend review of these algorithms by updating them according to the current medical literature so as to reduce strain on local ED resources. We also suggest that a larger, multi-center study is needed to prospectively analyze the appropriateness of ED visits once these revisions have been instituted.