Objective. To compare daytime nurse telephone triage calls received by a suburban practice with those received by a hospital-based, inner-city pediatric practice.
Methods. A research assistant, listening simultaneously with triage personnel, prospectively coded all calls received by the nurse triage telephone offices in 2 pediatric practices. Calls were coded the first full, nonholiday week of 3 consecutive summer and 3 consecutive winter months, alternating morning and afternoon sessions. One practice was suburban and had almost all commercially insured patients; the other was hospital-based, in an inner-city, and consisted mostly of patients with Medicaid coverage or no health insurance.
Results. A total of 901 calls were triaged in the suburban practice (SP) and 768 in the urban practice (UP). The chief complaints of calls regarding medical problems were similar at both sites. Difficulties with language were noted less often in SP compared with UP (1% vs 17%). The reason for the calls differed by site: medical problem relating to illness or injury (SP 55% vs UP 40%); social issue (SP 1% vs UP 9%); documentation request (SP 2% vs UP 7%); request for laboratory work (SP .3% vs UP 4%); and well child advice (SP 9% vs UP 5%). The disposition of calls also differed by site: telephone advice was offered significantly more in SP than in UP (32% vs 20%); fewer calls required the need for the medical record in SP than in UP (2% vs 12%).
Conclusions. Calls received by a daytime nurse telephone triage office in an affluent SP and a UP are similar in regard to medical problems. Training programs can feasibly prepare physicians and nurses for both kinds of practice settings. The urban site received more calls affected by language and social issues. This could have administrative implications for staffing ratios, language skills of staff and knowledge of available support services.
Telephone triage is an important component of pediatric practice and has traditionally been divided into daytime and after-hour service systems. Poole1 reviewed calls received over a 4-year period by the After Hours Program in Denver. Over 100 000 phone calls were managed successfully by trained nurses covering 92 physicians from 56 practices. In subsequent studies, safety, satisfaction and cost have been examined with no adverse outcomes reported.2–4 The set-up and management of a daytime nurse telephone triage system was recently reviewed by Schmitt.5,,6 However, in contrast to after-hour services, little is known about the content of daytime nurse telephone triage calls. The purpose of this study was to describe the content of calls received during daytime office hours at 2 telephone triage offices.
Daytime telephone calls triaged by nurses at 2 pediatric practices, one suburban and the other urban, were studied. The suburban practice (SP) was the Department of Pediatrics at Lahey Clinic in Burlington, Massachusetts. In 1997, this practice had 28 000 visits from 10 000 patients. More than 90% of the patients were commercially insured, the majority with health maintenance organization insurance. Six physicians worked as 5.9 full-time equivalents (FTE). The urban practice (UP) was the Pediatric Practice at Boston Medical Center (formerly Boston City Hospital). This practice was hospital-based and located in inner-city Boston. In 1997, 19 000 visits were seen from a pool of 8000 patients. More than 80% of patients were covered by Medicaid or were uninsured. The practice employed 10 pediatricians and 2 nurse practitioners working as 7 FTE. The UP served as a continuity site for pediatric residents, however, all calls at this practice were processed through the nurse telephone triage office.
Two research assistants (RA) were employed for the study; each was consistently assigned to the same practice. The RA listened to calls simultaneously with triage personnel and could not affect the length or disposition of the calls. Telephone triage was defined as telephone calls incoming to a nurse telephone triage line. Both triage lines were answered by a registered nurse with pediatric experience and neither practice charged for nurse telephone consultation. The calls were prospectively coded by the RA during the first full, nonholiday week of 3 consecutive summer and 3 consecutive winter months, alternating 4-hour morning and afternoon sessions (8am–12 pm and 1–5 pm).
The RA timed the length of the telephone call from the start to the completion of the call. Callers to the SP initially reached an automated phone message, in English, that allowed the caller to choose from a menu that included one option to make schedule changes and another option to speak to a nurse. Callers to the UP spoke directly to a clerk who may or may not have been fluent in languages other than English. The clerk triaged all calls of a medical nature to the nurse; the clerk did not schedule any same-day sick appointments. Parents did not directly page clinicians at either site.
Information coded included call data (date, time, length of call), caller and patient information (primary language of caller, relationship of caller to patient, patient age), reason for the contact, and disposition of call. A chief complaint was recorded for every contact code in the medical problem category (illness/injury, previously seen, patient with illness in emergency department, well-child care).
At both practices, medical record department staff managed all medical records, which were stored in a space located in the same building as the practice. Neither site had electronic medical records.
The language of the caller was coded as either English-speaking or non–English-speaking/language-impacted. The call was coded as language-impacted if, in the judgment of the RA listening, the caller spoke English but was difficult to understand or, if because of limited language skills, the caller could not find the correct word to express a key component of the exchange. Approximately half of the nurses at the urban site were fluent either in English and Spanish or English and French Creole. Many of the calls to the UP started in English and then switched to the caller's native language if the nurse also spoke that language.
A χ2 test was used to test the difference of proportions in the results.
Institutional review board approval was obtained at both sites.
A total of 1669 telephone calls were triaged, 901 calls in the SP and 768 calls in the UP. The majority of calls were received during the morning sessions at both practices. Differences emerged between sites with respect to identification of caller, length of call, and language of the caller (Table 1). In the SP, the mother was the caller more often. In the urban site, the 32% of the callers who were not the mother were distributed as follows: 9% other family member, 8% medical personnel, 5% emergency department, 3% WIC office, 7% other.
In the UP, the language of the caller was an issue more often, affecting 1 out of every 6 telephone encounters. To assess the impact of language on the length of the call, the length of the calls from non-English and language-impacted callers (grouped as language problem) were compared with English-speaking callers. Calls that were categorized as language problem lasted longer than those that were not categorized as such (>3 minutes, 51% vs 39%; P = .02).
The reason for the calls and the disposition of the calls differed by site (Table 2). More calls at the UP focused on social issues (1 out of every 11 calls), documentation requests, and laboratory requests while more calls to the SP were for well-child advice. With respect to disposition, more calls to the SP were handled with telephone advice compared with the UP. One out of every 8 calls to the UP resulted in the need for the medical record to be obtained.
The chief complaints of calls received about medical problems were similar at both sites (Table 3). Common chief complaints included cough with or without fever, sore throat with or without fever, ear complaint with or without fever, and rash with no fever.
Telephone triage is an important part of pediatric practice. Little is known about the content of daytime telephone triage calls. This is the first study we are aware of that has attempted to describe the content of daytime telephone triage calls and analyze variations that may exist between practices.
The chief complaints of calls received regarding medical problems were similar at both sites. They were also strikingly similar to the chief complaints received after office hours as reported by Poole.1 Common reasons for calling included cough, vomit/diarrhea, ear complaint, sore throat, rash, and trauma or injury. The data are helpful to guide curriculum for any practice wishing to provide educational training for nurse triage staff. The most frequent medical call to the suburban site concerned trauma or injury. We speculate that this was attributable to insurance restrictions that limited suburban patients from being seen in an emergency department without calling the practice first. We also believe that many patients of the UP were still in the habit of using the hospital emergency department, located one building away from the practice, without calling the practice first.
The challenges of language and social issues found more often in the urban triage office may have important implications for administrative operations. It has been a long-standing assumption among health professionals working in an inner-city or low socioeconomic site that it takes more time and resources to care for families in these settings. Our study supports this impression. This study's findings indicate caller language is a common problem in the urban setting. Improving communication between staff and non–English-speaking parents is costly, including hiring interpreter staff, recruiting bilingual staff or bringing the patient into the office, perhaps unnecessarily, for a same day appointment. Dealing with social issues also requires time and support staff. The need to retrieve the medical record and complete documentation requests increases administrative costs.
This study had a number of limitations. First, only 2 practices were studied, one in a suburban location and the other in the inner-city, arguably serving patients at opposite ends of the economic scale. Hence, generalizability may be limited, particularly for practices that care for patients from a wider socioeconomic range than either of these sites. We were careful in assessing calls in the morning as well as the afternoon and conducted the study during summer and winter months to ensure a representative sample of all calls to the practices. Second, the coding system was developed specifically for this project based on the experience of 2 of the principal investigators. It is possible that other valid coding systems exist. Although formal comparison of coding between the RAs was not done, regular meetings with the principal investigators and the RAs were conducted and formal definitions were provided for all coding categories (language, contact, and disposition). Third, 20% of the calls received by the UP were not timed because of the frequent disruptions in the triage office.
A nurse telephone triage office improves the quality of a pediatric practice. Access for families is improved. The knowledge and teaching skills of nursing staff are used appropriately and using nurses on the telephones frees up physician time for other issues.
This study was supported by the Joel and Barbara Alpert Children of the City Endowment Fund.
We thank the nursing staffs, led by Kay Cahn, RN, at Boston Medical Center and Jane Silveria, RN, at Lahey Clinic, for their cooperation with this study.
- Received March 31, 1999.
- Accepted February 15, 2000.
Reprint requests to (B.L.P.) Division of Pediatric Ambulatory Services, ACC 5, Boston Medical Center, 850 Harrison Ave, Boston, MA 02118. E-mail:
This was a poster presentation at the Pediatric Academic Meetings; May 1998; New Orleans, LA.
- SP =
- suburban practice •
- FTE =
- full-time equivalents •
- UP =
- urban practice •
- RA =
- research assistant
- Poole SR,
- Schmitt BD,
- Caruth T,
- Peterson-Smith A,
- Slusarski M
- ↵Kempe A, Poole SR, Dempsey CL, Bothner J, Schmitt BD. Appropriateness of referral for after hours evaluation using an automated pediatric telephone triage system. Ambulatory Child Health. 1997;3:140. (Abstract 25)
- ↵Frei N, Kempe A, Leamer KA, Schmitt BD, Poole SR, Hegarty TW. Reducing after hour referrals from an automated telephone triage system with second level physician triage. Ambulatory Child Health. 1997;3:140. (Abstract 26)
- ↵Poole S, Kempe A, Hegarty TW, Schmitt BD. Evaluation of parent or nurse override decisions using automated telephone triage and advice algorithms. Ambulatory Child Health. 1997;3:139. (Abstract 21)
- Schmitt BD
- Schmitt BD
- Copyright © 2000 American Academy of Pediatrics