Table 2 Effect of interventions by population-intervention-resource use combination

Decreased subsequent healthcare resource use

Frequent Attenders

–

ED revisit

Care plans

Case management

Social work home visits

Diversion strategies to nonurgent care

Printout case notes

Medical Care Plan**

Care Co-ordination**

Disease Management**

In ED frequent attendance patients, interventions, decreased ED revisits. This is based on high confidence data from 3 reviews

(Moe et al. 2017) Median rate ratio was 0.63 (IQR = 0.41 to 0.71), general effect of interventions was to decrease ED visits post-intervention. Data from 10/31 primary studies

AMSTAR II of review = High. GRADE of outcome = not reported. RoB = 7 Moderate, 3 High

(Berkman et al. 2021) – Reduction in ED revisit

3/4 RCT = reduction, 1/4 RCT = no difference, 1/2 OBS = reduction. 1/2 OBS samples = reduction one control group and no difference with one control group

AMSTAR II of review = High. GRADE of outcome = Moderate

RoB of primary studies = Low = 1, Some Concerns = 3, High = 2

(Althaus et al. 2011)– 7/11 decrease primary studies, 1 increase primary study, 2 not reported

AMSTAR II of review = High. GRADE of outcome = not reported. RoB = reported individually see online supplement

This data is supported by 1 Moderate confidence reviews (Wong et al. 2020). Reduction in visits between 48.4 and 89.5%

GRADE = not reported. RoB/Quality, 2 = moderate quality 3 = low quality primary studies

Supported by 3 Critically Low confidence reviews ((Iovan et al. 2020), (Mauro et al. 2019), (Deschamps et al. 2021))

11/17 decrease primary studies, 7/17 No effect (1 study reported twice),

2/4 decrease primary studies, 1/4 No effect, 1/4 Unable to comment

2/5 decrease primary studies, 3/5 Unable to comment

Shared decision-making

–

ED Revisit

Provision of pre-test probability

In patients presenting with chest pain, interventions, decreased 7-day ED revisit rate. This is based on high confidence data from 1 review

(Flynn et al. 2012b), 1/1 decrease primary study (RCT)

AMSTAR II of review = High. GRADE of outcome = NR. RoB = Low, Quality of primary study = High

Alcohol

–

ED revisits

Screening and brief interventions

In patients screened for alcohol, screening and brief interventions, decreased ED revisits. This is based on critically low confidence data from 1 review

(Bray, Cowell and Hinde, 2011b), 3/4 decrease primary studies. 1/4 unable to comment

AMSTAR II of review = Critically low. GRADE of outcome = NR. Qualitative Methodological Scores = 13, 13, 14 (high)

Frequent Attenders

–

Inpatient admissions

ED- initiated patient navigation program

Emergency Room Decision- Support

Medical Care Plan**

Care Co-ordination**

Disease Management**

In ED frequent attendance patients, interventions, decreased inpatient admissions. This is based on high confidence data from 1 review

(Berkman et al. 2021)—Effect = Decrease, based on 1 × RCT, 1 × OBS study

AMSTAR II of review = High. GRADE of outcome = Low. RoB = RCT – Low, OBS – Some Concerns

This data is supported by Critically Low confidence data from (Iovan et al. 2020)—9 no effect, 9 decrease

Older Adults who fell

-

Hospital Admission

Interdisciplinary team

(Harper et al. 2021), RR 0.76; 95% CI 0.64–0.90,

AMSTAR II of review = High. GRADE of outcome = Moderate. RoB = Moderate to strong quality (RoB assessment included)

Short Stay Units in the ED

–

Hospital Admissions

ED short stay units

For patients in the ED, ED short stay units had a decreased Hospital admission. This is based on high confidence data from 1 review

(Galipeau et al. 2015)—3/3 primary studies positive

AMSTAR II of review = High. GRADE of outcome = low. RoB = Moderate

Shared decision-making

–

Testing

Chest pain decision aid

Provision of pre-test probability

In patients presenting with chest pain, interventions, decreased testing. This is based on high confidence data from 1 review

(Flynn et al. 2012b), 1/1 primary study for decreased cardiac testing at 30 days (decision aid), 1/1 positive primary study for decreased thoracic imagine (pre-test probability)

AMSTAR II of review = High. GRADE of outcome = NR. RoB = Low × 2, Quality of primary study = High × 2

Frequent Attenders

–

Cost

Care plans

Case management

Social work home visits

Diversion strategies to nonurgent care

Care coordination and community health worker program

Emergency Room Decision- Support (ERDS) program

Pain protocol

Individual Care Plan**

In ED frequent attendance patients, interventions, have a decreased Healthcare Costs. This is based on high confidence data from 2 reviews

(Moe et al. 2017)—11 decrease (RoB = 4 Moderate, 7 High)., 1 increase (outpatient costs, RoB = moderate), 1 no effect (non-ED costs, RoB = High)

AMSTAR II of review = High. GRADE of outcome = NR

(Berkman et al. 2021)—Effect = decrease, based on 2 of 3 RCT samples had favourable findings (RoB = 2 × Some Concerns), 1 of 3 RCT samples found no difference (RoB = 1 × Some Concerns), 1 of 1 OBS sample found no difference (RoB = 1 × Some Concerns)

AMSTAR II of review = High. GRADE of outcome = Low

This data is supported by 1 Moderate confidence review (Wong et al. 2020) which reported a reduction in costs in 3 studies

GRADE = not reported. RoB/Quality. 1 = moderate quality 2 = low quality

1 Critically Low confidence reviews (Mauro et al. 2019) reported a reduction in cost in 1 primary study (CASP Quality score 11/11)

Mixed effect on resource use

Lower Back Pain

–

ED revisit

MDT protocols

Clinical decision support

In patients with lower back pain, interventions, had a decrease to no effect on ED revisits. This is based on moderation confidence data from 1 review

(Liu et al. 2018)– 2 before and after primary studies. MDT protocols aimed at decreasing imaging for lower back pain decreased ED revisits, whilst clinical decision support had no effect

AMSTAR II of review = Moderate. GRADE of outcome = NR. Before and After Quality Assessment = low × 2

Older Adults who Fell

-

Hospital Admission

Interdisciplinary team

A non-significant reduction (P = 0.07) with intervention (RR 0.85; 95% CI 0.72–1.01, I2 0%). Heterogeneity: Tau(2) = 0.00, CHI(2) = 2.13, df = 4, test for overall effect 1.92 (p = 0.06)

Mental Health (acute suicidal ideation)

–

Psychiatric Admissions

Active follow-up and contact interventions

In patients presenting with acute suicidal ideation, interventions, both decreased and had no effect on Psychiatric Admissions. This is based on low confidence data from 1 review

(Inagaki et al. 2019), For psychiatric admissions at 12 months, 1 study had a decrease effect (RoB L = 3 U = 2 H = 2). 2 studies had no effect (RoB L = 4 U = 3 H = 0, L = 4 U = 1 H = 2)

AMSTAR II of review = Low. GRADE of outcome = NR. RoB = see above

Short Stay Units

–

ED revisit/Hospital Readmission

ED short stay units

For patients in the ED, ED short stay units had a decrease to no effect on ED revisit/Hospital readmission. This is based on high confidence data from 1 review (Galipeau et al. 2015)—2/4 decrease primary studies, 2/4 no effect primary studies

AMSTAR II of review = High. GRADE of outcome = low. RoB = Moderate

Increase in scheduled follow-up

Frequent Attenders

–

Primary Care Attendance

Patient navigation for ED patients

Emergency Room Decision- Support (ERDS) program

In ED frequent attendance patients, interventions, increased Primary Care visits. This is based on high confidence data from 1 review

(Berkman et al. 2021) – Effect = Increase, based on 1xRCT, 1 × OBS study

AMSTAR II of review = High. GRADE of outcome = Low. RoB = RCT – Low, OBS – Some Concerns

Asthma

–

Primary care follow-up

Educational interventions:

Arranged follow-up

Follow-up phone calls

Faxed letters

Oral steroids

Asthma action plans

In asthma patients, educational interventions increased scheduled follow-up rates with Primary Care Practitioners (aim of interventions). This is based on moderate confidence data from 1 study (Villa—Roel et al. 2016)

Risk Ratio = 1.6; 95% CI 1.31 to 1.87

AMSTAR II of review = Moderate. GRADE of outcome = NR. RoB = Unclear

Adults in the ED

–

Follow-up with specialist or primary care providers

Care Transition Interventions defined as:

Educational support (face-to-face, video-based or telephonic)

Reminders (mailed, text or telephonic)

Appointment scheduling

ED-based discharge instructions

Case management programs

In ED adult patients, care transition interventions, improve the rate of follow-up with specialist or primary care providers. This is based on high confidence data from 1 review (Aghajafari et al. 2020)

20 studies (8178 patients). ED-based CTIs increased odds of follow-up versus usual care (OR 1.79, 95% CI 1.43,2.24) AMSTAR II of review = High. GRADE of outcome = Low. RoB of primary studies = Low in 11/40

This data is supported by 1 critically low confidence review (Katz et al. 2012)

Based on 5/5 studies that demonstrated increased follow up (3/5 positive, 1/5 no effect, 1/5 NR). Grade and RoB not reported. JADAD score 3/5 and 2/5 in the two

Randomised studies

Frequent Attenders

–

Outpatient visits

Care plans

Case management

Social work home visits

Diversion strategies to nonurgent care

Printout case notes

In ED frequent attendance patients, interventions, have increased outpatient visits. This is based on high confidence data from 2 reviews

(Althaus F. et al. 2010)—Effect = 2 studies increased outpatient visits, as per aim

AMSTAR II of review = High. GRADE of outcome = not reported. Quality Criteria for NCBA studies = "Y = 6 U = 2 N = 2, Y = 7 U = 1 N = 2"

(Moe et al. 2017)—Effect = 6 studies increased outpatient attendances (RoB = 4 × Moderate, 2 × High), 2 no effect (RoB = 2 × High), 1 decreased outpatient attendances (RoB = 1 × Moderate). The aim of interventions was not reported

AMSTAR II of review = High. GRADE of outcome = not reported

*(Unable to establish if this was the aim of interventions in Moe et al.)

Lower Back Pain

–

Physio & Rehab

MDT protocols

In patients with lower back pain, interventions, increased the use of physiotherapy and rehabilitation services visits

This is based on moderate confidence data from 1 review

(Liu et al. 2018) – 1 before and after primary studies. MDT protocols aimed at decreasing imaging for lower back pain increased use of services

AMSTAR II of review = Moderate. GRADE of outcome = NR. Before and After Quality Assessment = low × 1

Risky Behaviour (Domestic Violence)

–

Increased Referral

Patient and physician notification

In patients presenting with a domestic violence related issue, interventions, increased referrals to services. This is based on low confidence data from 1 review

(Choo et al. 2012), 1/2 increase primary studies. 1/2 unable to comment

AMSTAR II of review = low. GRADE of outcome = NR. Quality of primary study = moderate

No effect on subsequent healthcare resource use

Older Adults –

ED revisits

Case management

Discharge planning

Complex geriatric assessment

In Older Adult ED patients, interventions had no effects on ED revisits. This is based on high confidence data from 4 reviews

(Hughes et al. 2019), RR = 1.13; 95% CI 0.94—1.36

AMSTAR II of review = High. GRADE of outcome = high. RoB of primary studies = Low to High

(Harper et al. 2021), RR 0.85; 95% CI 0.72–1.01

AMSTAR II of review = High. GRADE of outcome = Low. Quality of primary studies = Moderate to Strong (RoB assessment included)

(Hesselink, Sir and Schoon, 2019), 1/4 Primary study positive effect at 1 and 3 months, 4/4 Primary studies = No effect,

AMSTAR II of review = High. GRADE of outcome = NR. RoB of primary studies = Moderate to high

(Elliott et al. 2022), 5/8 primary studies = Positive effect, 3/8 = No effect

AMSTAR II of review = High. GRADE of outcome = NR. RoB = Low to Moderate (only in randomised trials)

This data is supported by 3 Moderate confidence reviews ((Galvin et al. 2017), (Santosaputri E., Laver K., and To T., 2019), Cassarino), 2 Low confidence reviews

((Fealy et al. 2009), (Ratsimbazafy et al. 2020)) and 3 Critically Low confidence reviews ((Sinha et al. 2011), (Aminzadeh and Dalziel, 2002), (Karam et al. 2015))

5/10 decrease primary studies, 4/10 No effect primary studies, 1/10, Unable to comment primary studies

Adults in the ED

–

ED revisits

Care Transition Interventions defined as:

Educational support (face-to-face, video-based or telephonic)

Reminders (mailed, text or telephonic)

Appointment scheduling

ED-based discharge instructions

Case management programs

In ED adult patients, care transition interventions, have no effect on ED revisits. This is based on high confidence data from 1 review (Aghajafari et al. 2020)

20 studies (8048 patients). ED-based CTIs had no effect on ED revisit (OR 1.01, 95% CI 0.86, 1.20), (experimental group events = n = 845, control group events = n = 832)

AMSTAR II of review = High. GRADE of outcome = Low. RoB of primary studies = Low in 12/20

This data is supported by 1 Critically low confidence review (Katz et al. 2012). Based on 3/5 (1xRCT, 2xOBS) studies that demonstrated no effect on ED revisits. Grade and RoB not reported. JADAD score 3/5 RCT

In adult patients, telemedicine interventions had no effect on ED revisit based on Critically low narrative data from (Hersh et al. 2001), based on 1 RCT. Grade and RoB not reported

Asthma

–

ED revisit

Educational interventions:

Arranged follow-up

Follow-up phone calls

Faxed letters

ral steroids and transport vouchers

Asthma action plans

In asthma patients, educational interventions, had no effect on Asthma relapses (including ED revisits). This is based on moderate confidence data from 1 study

(Villa—Roel et al. 2016)) – Risk Ratio = 1.3 (95% CI 0.82 to 1.98)

AMSTAR II of review = Moderate. GRADE of outcome = NR. RoB = Unclear

This data is supported by 1 moderate confidence narrative review (Villa-Roel et al. 2018) based on one high RoB and one low RoB studies (one had a decrease effect, the other an increase effect for AAP and % relapses)

This evidence is supported 1 low confidence review. (Tapp, Lasserson and Rowe, 2007) No effect on ED revisit. Grade of outcome = low. Based on three RCTs with mixed RoB

Antibiotics

–

ED revisit

Pharmacist lead algorithm

Pharmacist culture follow-up

Pharmacist presence

In ED patients, pharmacist interventions, had no effects on ED revisits. This is based on high confidence meta-analysis data from 1 review

(Kooda, Canterbury and Bellolio, 2022) OR of 0.65 (95% CI 0.39 to 1.10) (Tau2 = 0.42, CHI2 = 53.57, df = 9 P < 0.00001, I2 = 83%, Z = 1.59 p = 0.11)

AMSTAR II of review = High. GRADE of outcome = NR. Newcastle–Ottawa RoB

Moderate 9/10, High 1/10, NIH Quality Score Fair = 7/10, Good 3/10

This data is supported by 1 Low confidence review (Losier et al. 2017). 1 study (high RoB) demonstrated a decrease effect, 1 study (high RoB) demonstrated a positive effect on ED revisit

General Practitioners in the ED

–

ED revisit

GPs in the ED

For patients in the ED, being seen by a GP had no effect on ED Revisits. This is based on high confidence data from 1 review

(Gonçalves-Bradley et al. 2018)—1 primary study. 17% (95% CI 15.7% to 18.8%) of patients seen by a GP, and 18% (95% CI 16.3% to 19.5%) of patients seen by an Emergency Physician re-attending the ED for the same problem within 30 days of index visit

AMSTAR II of review = High. GRADE of outcome = very low. RoB = L = 3 U = 8 H = 3

Adults with chest pain

–

ED revisit

CCTA

For chest pain, CCTA had no effect on ED revisit. This is based on low confidence data from 1 review. (Hulten Edward et al. 2013)

Pooled weighted odds ratio (range) 0.94 (0.67–1.31, p 0.70) I2 = 0.0%, p = 0.68

AMSTAR II of review = Low. GRADE of outcome = NR. RoB = Low-Unclear

Mental Health (acute suicidal ideation)

–

ED Contacts

Active follow-up and contact interventions

In patients presenting with acute suicidal ideation, interventions, had a no effect on ED contacts. This is based on low confidence data from 1 review

(Inagaki et al. 2019), 1 primary study showed no effect

AMSTAR II of review = Low. GRADE of outcome = NR. RoB = L = 4 U = 1 H = 2

Mental Health (acute suicidal ideation)

–

GP Contacts

Active follow-up and contact interventions

In patients presenting with acute suicidal ideation, interventions, had a no effect on GP contacts. This is based on low confidence data from 1 review

(Inagaki et al. 2019), active contact resulted in a reduction at 3 months but this was reversed to an increase at 12 months (n = 1, RoB L = 4 U = 3 H = 0). 2 other studies (presented in 3 papers) showed no effect (n = 3 RoB L = 5 U = 1 H = 1, L = 5 U = 1 H = 1, L = 4 U = 2 H = 1). AMSTAR II of review = Low. GRADE of outcome = NR. RoB = see above

General Practitioners in the ED

–

GP visits

GPs in the ED

For patients in the ED, being seen by a GP had no effect on GP visits. This is based on high confidence data from 1 review

(Goncalves-Bradley D. et al. 2018)—2 primary studies. No effect

AMSTAR II of review = High. GRADE of outcome = very low. RoB = L = 3 U = 8 H = 3, L = 5 U = 4 H = 5

Mental Health (acute suicidal ideation)

–

Psychiatric Contacts

Active follow-up and contact interventions

In patients presenting with acute suicidal ideation, interventions, both decreased and increased on Psychiatric contacts. This is based on low confidence data from 1 review

(Inagaki et al. 2019), For psychiatric contacts at 12 months, 1 study had a decrease effect (L = 1 U = 1 H = 4), 1 had an increase effect (RoB L = 5 U = 1 H = 1)

AMSTAR II of review = Low. GRADE of outcome = NR. RoB = see above

Older Adults

–

Hospital re-admissions

Case management

Discharge planning

Complex geriatric assessment

In Older Adult ED patients, interventions had no effect on Hospital re-admissions. This is based on high confidence data from 2 reviews

(Hughes et al. 2019), Relative risk [RR] = 0.96; 95% CI 0.51–1.83

AMSTAR II of review = High. GRADE of outcome = Low. RoB of primary studies = Low to High

(Elliott et al. 2022), 2/2 primary studies = No effect

AMSTAR II of review = High. GRADE of outcome = NR. RoB = Low 1/2, NR in 1/2

This data is supported by 2 Moderate confidence reviews ((Cassarino et al. 2019; Santosaputri E., Laver K., and To T., 2019)) and 2 Low confidence reviews ((Ratsimbazafy et al. 2020; Fealy et al. 2009))

4/7 decrease primary studies, 3/7 No effect primary studies

Adults with chest pain

–

Hospital Admission

CCTA

For chest pain, CCTA had no effect on hospital admissions. This is based on low confidence data from 1 review. (Hulten Edward et al. 2013)

Pooled weighted odds ratio (range) 1.20 (0.67–2.16, p 0.50) I2 = 0.0%, p = 0.68)

AMSTAR II of review = Low. GRADE of outcome = NR. RoB = Low- Unclear

Adults in the ED

–

Hospital Re-admission

Care Transition Interventions defined as:

Educational support (face-to-face, video-based or telephonic)

Reminders (mailed, text or telephonic)

Appointment scheduling

ED-based discharge instructions -Case management programs

In ED adult patients, care transition interventions, had no effect on hospital re-admissions. This is based on high confidence data from 1 review. (Aghajafari et al. 2020)

13 studies (5742 patients). ED-based CTIs had no effect on hospital admissions (OR 0.99, 95% CI 0.86,1.14)

AMSTAR II of review = High. GRADE of outcome = Low. RoB of primary studies = Low in 11/40

This data is supported by a 1 Critically low confidence review ((Katz et al. 2012)). Based on 1/5 (1xOBS) studies that demonstrated increased hospitalisations. Grade and RoB not reported

Unclear aim of intervention/not possible to evaluate

Alcohol

–

Outpatient Resource Use

Screening and brief interventions

In patients screened for alcohol, screening and brief interventions, increased outpatient resource use. This is based on critically low confidence data from 1 review

(Bray, Cowell and Hinde, 2011b), 2/4 increased resource use. 2/4 unable to comment

AMSTAR II of review = Critically low. GRADE of outcome = NR. Qualitative Methodological Scores = 13, 12 (high)

*(Unable to establish if this was the aim of interventions)

Palliative Care

–

ED revisit

N/A

From 1 review it is not possible to comment on the effect of Palliative Care ED interventions on subsequent healthcare resource use. (da Silva Soares, Nunes and Gomes, 2016)