Nurse Telephone Triage in Primary Care: ESTEEM Trial Analysis

Introduction

Nurse-led telephone triage has increasingly been employed to manage physician consultation demands in British general practice. Prior research is vague when it comes to delineating the association between nurse triage call outcomes and practical clinical experience. A majority of studies are restricted to scrutinizing out-of-hours nurse phone triage. This study was conducted within the framework of current knowledge, thereby contributing a solution angle highly relevant to the nursing profession. The study's key purpose was to investigate whether primary care nursing staff who undertake digital decision-supported software phone triage possess professional traits that are associated with call disposition (Varley et al., 2016).

Telephone triage is an approach utilized for addressing increasing nurse workloads. Nursing staff offer an effective and safe triage service across various settings, including primary care, where they may prove efficient in managing physicians' workloads in both out-of-hours and same-day primary healthcare facilities.

Literature Review and Theoretical Framework

Qualification levels, duration of experience, and other nursing traits can impact triage outcomes. An analysis of 60,794 calls handled by 296 National Health Service Direct nurses revealed a positive link between call disposal trends and duration of experience. Relatively new nursing staff — those with fewer than ten years in the field — were less likely to dispose of calls as self-care recommendations compared to nursing staff practicing for two decades or more. This finding aligns with other studies indicating that more experienced nursing staff exhibit greater accuracy in selected triage results, lower data acquisition levels during triage evaluations, and more inference and judgment-making based on past experience.

Research on nursing characteristics represents a continuous endeavor carried out over many years. It appears that numerous studies published several years ago continue to apply in the present context. The researchers referenced journal articles published between 1995 and 2016.

This research did not directly apply formal theoretical concepts. Instead, it applied general nursing-related research hypotheses and problems. Questionnaires administered to nursing staff delivering the ESTEEM trial's nurse intervention branch effectively captured type of role (i.e., nurse practitioner or practice nurse), duration of experience, prescriber status, educational qualification, perceived triage readiness, and past triage experience. The key outcome sought was the share of triaged nursing care patients who received a follow-up recommendation within the practice — that is, call disposition — encompassing all forms of contact (in-person, home visits, and telephone) by nurses or physicians.

The study incorporated nursing theory elements relevant to the research problem. Every nurse participated in a customized training initiative to facilitate their use of the clinical decision support software (CDSS) provided by Plain Healthcare during phone triage consultations. All study subjects potentially engaged in a triage role were expected to receive the standardized training package relating to phone consultation skills and CDSS application. It is important to note that this support software was designed to assist nursing professionals in care-related decision-making. While nurses were required to log in to the software for all consultations, they were not obligated to follow the software-generated recommendation during the trial. The software's intent was therefore to facilitate nurse decision-making while offering flexibility for applying their expertise and experience in determining triage call outcomes.

The stepwise strategy employed in this study proceeded as follows:

Variables, Hypotheses, and Research Design

1. Extraction of patient information from clinician forms (primary result modeling); 2. Identification of likely explanatory factors from clinician forms; 3. Survey of nursing staff; and 4. Data acquisition at the practice level.

Covariates at the patient level were: participant sex (reference: female); age categorized as 0–4 years, 5–11 years, 16–24 years, 25–59 years (reference category), 60–74 years, and 75+ years; duration of experience; deprivation at the patient level segregated into rank-based quintiles; and ethnicity.

Covariates at the practice level were: location; deprivation at the practice level (non-deprived or deprived; below average or average deprivation for England as the reference group; and above-average England-based deprivation); and size of practice list (small: fewer than 3,500 registered patients; large: more than 8,000 registered patients, serving as the reference category; and medium: between 3,500 and 8,000 registered patients).

All variables listed above are independent. The study establishes call disposition as the dependent variable. Suggested call disposition was the chief outcome, dichotomized as a binary variable indicating whether practice follow-up was needed or not. All statistical analyses were composed of hierarchical logistic models. Distinct patient examinations were nested within nurses and vice versa; therefore, both practice and nurse are treated as random effects in the study (Varley et al., 2016).

Variable definitions, levels, and context were not always clearly delineated. However, the researchers noted a general survey parameter application.

The linkage between triage call disposition and nurse traits was examined for same-day patient visit requests within general practice by employing multivariable logistic regression modeling. Discrete nurse traits were studied for their association with follow-up recommendations. A model encompassing all nursing traits was also constructed. All nurse traits found to be significant — individually or within the comprehensive nurse trait model — were included in further models.