Table 1

Characteristics of included studies (31 studies)

Main outcome areaAuthor
year
Country
Reference		Study designSample/data sizeUrgent or emergency careStaff type conducting triageParticipants and service nameComparatorQuality
User experienceBjörkman
2018
Sweden44		Qualitative:
'Netnographic' method using information from online forums using six step		Data collected from 3 online forumsUrgentNurseGeneral populationNoneHigh
User experienceO'Cathain
2014
England40		Quantitative:
Survey		Survey sent to 1200 patients from 4 pilot sites, 1769 responded and were included for analysisUrgentNon-clinical call handlerGeneral populationNoneMedium
User experienceMcAteer
2016
Scotland6		Mixed methods: survey and interviewsSurvey: Age and sex-stratified random sample of 256 adults from each of 14 Scottish GP surgeries, final sample was 1190.
Interviews: 30 semistructured interviews		UrgentNon-clinical call handlerGeneral population (National Health Service (NHS) 24 users and non-users)Interviewees (from survey respondents) grouped into satisfied users, dissatisfied users and non-usersHigh
User experienceRahmqvist
2011
Sweden41		Quantitative:
Survey		Random sample of 660 callers, made at one call centre site in October 2008UrgentNurseGeneral population(1) Cases: those who disagreed with nurse advice and felt they needed higher level of care; (2) Controls: those who disagreed with nurse advice OR felt they needed higher level of care; (3) other callersMedium
User experienceGoode
2004
England43		Qualitative:
Interview study		60 interviewsUrgentNurseGeneral populationNoneHigh
User experienceWinneby
2014
Sweden45		Qualitative:
Interview study		8 semistructured interviewsUrgentNurseGeneral populationNoneHigh
User experienceGoode
2004
England42		Qualitative:
Interview study		10 semistructured interviewsUrgentNurseInterviews focused on menNoneHigh
Patterns of triage advicePayne
2001
England23		Routine data analysis56 450 callsUrgentNurseGeneral populationNone—comparisons within digital triage call dataHigh
Patterns of triage adviceElliot
2015
Scotland5		Routine data analysis1 285 038 callsUrgentNurseGeneral populationNone—comparisons within digital triage call dataHigh
Patterns of triage adviceZwaanswijk
2015
Netherlands25		Routine data analysis895 253 patientsUrgentNurse (GP cooperative)General populationSome comparison with non-digital triageHigh
Patterns of triage adviceNjeru
2017
USA7		Routine data analysis587 cases
587 controls		UrgentNurseThose aged over 18— (callers with and without limited English proficiency)Patients with limited English proficiency compared with English proficientHigh
Patterns of triage adviceJacome
2018
Portugal24		Routine data analysis148 099 callsUrgentNurseGeneral population
(Older age groups 65+)		None - Comparisons within digital triage call dataHigh
Patterns of triage adviceHsu
2011
England21		Routine data analysis402 959 callsUrgentNurseOlder age groups (aged over 65 years)NoneHigh
Patterns of triage adviceCook
2013
England20		Routine data analysis358 503 callsUrgentNursechildren aged 0–15
(<1, 1–3 and 4–15 years))		Comparisons between age groupsMedium
Patterns of triage adviceNorth
2010
USA22		Routine data analysis20 230 callsUrgentNurseGeneral population (those with subscription and insurance)Three comparison groups:
(1)Triaged callers; (2) Emergency Department (ED) attendances; (3)
Office (GP) visits. (Comparison of hospitalisation in these groups)		Medium
Patterns of triage adviceNorth
2011
USA19		Routine data analysisOver the 3-year period: 105 866 adult calls (65% of the total calls). Of these, 14 646 (14%) were made by a surrogate on behalf of the patient.UrgentNurseGeneral population (aged over 18)Surrogate vs self callsMedium
Service use following triageLattimer
2000
England32		Quantitative descriptive: Cost-effectiveness report from controlled trial>14 000
Control group (n=7308 calls)
Intervention group that is, Nurse telephone consultation (n=7184 calls)		UrgentNurse (within general practice cooperative)General populationUsual care (referral to a General Practice) compared with
nurse-led digital triage		Medium
Service use following triageMunro
2000
England29		Routine data analysisStudy corresponds to the 1st year of operation, where 68 500 NHS direct calls from the 1.3 million people served.UrgentNurseAll contacts with these immediate care services (at time spanning before and after introduction of call centre based service)Service use in regions where digital triage service was introduced, compared with regions with no implementationHigh
Service use following triageDale
2003
England36		Controlled trial635 triaged calls
611 non-triaged calls		EmergencyNurse and paramedic (within emergency control room)General population, calling the emergency service for non-emergency concerns (only those aged 2+)The control group not offered triage was compared with calls digitally triaged either by nurses or paramedics.High
Service use following triageFoster
2003
England27		Routine data analysis and data linkage4493 calls, of which 193 were advised to go to Emergency Department (ED)UrgentNurseGeneral populationThree comparison groups:

  1. Callers triaged to Emergency Department (ED), who attended

  2. Callers triaged to ED who did not attend

  3. Callers with different triage outcome who attended ED.

Medium
Service use following triageMark
2003
England46		Mixed methods (routine data analysis +interviews)Numbers of calls analysed across 3 years:
5126 (year 1998)
5702 (1999)
4698 (2000)		UrgentNurseGeneral populationn/aLow
Service use following triageSprivulis
2004
Australia34		Routine data analysis & data linkage13 019 presentations to Emergency Department (ED) of which 842 were identified as having contacted Health-Direct within the 24 hours period prior to presentation.UrgentNurseGeneral population—all patients who contacted the digital triage service during the 1-year study period

  1. Patients who were digitally triaged prior to attending ED

  2. Patients who were not digitally triaged

High
Service use following triageDunt
2005
Australia30		Quantitative: four trials including surveys (self-reported service use)Random sampling (350 households per trial site)UrgentNurseGeneral population2 sites using ‘standalone’ telephone triage which used ‘call centre software’
2 embedded telephone triage sites using paper based protocols		Medium
Service use following triageMunro
2005
England28		Quantitative: Surveys (care providers)571 surveys sent (188/297) responses from GP cooperatives, (35/35) for ambulance services and (200/239) for emergency departmentsUrgentNurseSurveys sent to care providers (general use of services following NHS direct implementations)n/aMedium
Service use following triageStewart
2006
England37		Routine data analysis & data linkage3312 calls to call centre based service, and 14 029 patients who attended Emergency Department (ED)UrgentNurseChildren and young adults aged under 16

  1. Patients advised through digital triage to attend ED

  2. Patients given alternative referral advice, through digital triage, but who still attended ED

  3. Patients referred to ED by their GP

  4. Self-referrals to ED

High
Service use following triageByrne
2007
England26		Quantitative: Survey268 callersUrgentNurseGeneral public with 3 symptom types (abdominal pain or cough and/or sore throat)NoneHigh
Service use following triageMorimura
2010
Japan35		Routine data analysis26 138 telephone consultationsEmergencyNurse and call handlerGeneral populationNoneMedium
Service use following triageHuibers
2013
Netherlands33		Quantitative:
Questionnaires		7039 questionnaires returned (from a total of 13 953 sent)UrgentNurseGeneral population (users who had a telephone contact with a nurse)NoneHigh
Service use following triageTurner
2013
England38		Routine data analysis400 000 calls to call centre based service in first year of operation analysedUrgentNurseGeneral populationMatched sites: (1) Intervention sites: four digital pilot sites;
(2)Control sites (North of Tyne, Leicester, Norfolk)		High
Service use following triageTurbitt
2015
Australia31		Quantitative:
Surveys		1150 parents attending Emergency Department (ED) (decline rate 19.9%)UrgentNurseSpecific groupSome comparisons between parents who called and did not call but prior to attending EDMedium
Service use following triageSiddiqui
2019
Australia39		Routine data analysis and data linkage12 741 triaged cases linked to 72.577 ED presentationsUrgentNurseGeneral populationNoneHigh

  • ED, emergency department; GP, general practice.