Table 2 Characteristics and outcome of included quasi-experimental studies

Le Menach, 2014 [60]

Children aged 6 months – 16 years from one general practice.

B&A

MMR

Multi-component

Campaign offering accelerated vaccination (6-11 months), early 2nd dose (6-11 months) and catch up vaccinations.

Coverage in 1538 children measured before and after campaign.

Increase in proportion of >14 months immunised by 3% (to 71%) and of >60 months by 5% (to 65%) following the campaign.

Low (assessors not blinded; unclear consideration of group effects)

Cockman, 2011 [61]

All children in the London Borough of Tower Hamlets.

Eco

MMR

Multi-component

Quality improvement project associated with campaign including: incentive payments; practice network; commissioning care package; new targets; IT for reminder/recall; active follow-up defaulters.

Coverage in all children in the area measured over time.

Coverage of MMR1 increased from 80% before the intervention to 94% after. Significant difference (p=0.001) in slope of coverage change pre (-0.07% per quarter) and post (1.86% per quarter).

Low (assessors not blinded; unclear consideration of group effects)

Siriwardena, 2003 [57]

Selected general practices from all practices in Lincolnshire.

B&A

Influenza & pneumococcal

Multi-component

Dissemination of clinical guidelines; advise on data and surveillance; organisational strategy; reminder/recall; comparative performance.

Coverage in 21 general practices before and after participating in project.

Significant increases in coverage before and after for both influenza and pneumococcal in a range of groups e.g. pneumococcal in CHD 27.5% increase (CI 12.6-42.3%; p=0.002)

Moderate (general objective; diffuse intervention; assessors not blinded; no consideration of group effects)

McDonald, 1997 [66]

Eligible patients in risk groups registered at general practices in Tameside.

B&A

Pneumococcal

Multi-component

Improved vaccine supply; clinical guidelines; patient materials; patient information leaflet translations; education.

Proportion of eligible patients immunised at participating practices before and after the intervention.

Increase in coverage from 6% before to 33% after the campaign.

Moderate (eligibility and selection unclear; diffuse intervention; assessors not blinded; no consideration of group effects)

MacDonald, 2016 [62]

Unimmunised children from Dudley local health area.

Eco ITS

MMR

Outreach

Immunisation offered during home visits

Comparison of coverage in local population using quarterly routine data.

Intervention contributed 2.6% of the MMR doses given during the study period.

High (enrolment and sample size unclear; assessors not blinded; limited statistical consideration)

Atchison, 2013 [63]

General practices in Wandsworth, London.

B&A

Childhood schedule

Reminder/recall

Standardised reminder/recall system involving letters and referral to health visitors.

32 participating practices compared to 44 not participating before and after the intervention.

Significant increase in coverage in intervention group, but as a result of unexplained decreases in control group coverage.

Moderate (likely differences between intervention and control practices; assessors not blinded; no consideration of group effects).

Henderson, 2004 [64]

General practices in Highland NHS Health Board area.

Eco

Childhood schedule

Reminder/ recall

Participation in national reminder/recall system vs. use of general practices’ own system.

Coverage between 8 practices using their own reminder/recall system vs. 66 participating in a national system.

Higher coverage in national system practices of diphtheria by age 2 (6.4%, CI 1.7-11.1, p=0.001) and Men C by age 2 (7.6%, 2.7-13%, p=0.001) but no difference for diphtheria by 1 year or MMR by 2 years.

Moderate (natural experiment – intervention dose unclear; likely difference between intervention and control practices; assessors not blinded; group effects not considered)

Gosden, 2003 [65]

Selected general practices in England.

B&A

Childhood schedule

Incentive

General practice contracting arrangements: GMS vs. PMS contracts.

Coverage in 10 practices who had switched to PMS contract vs. matched 10 control practices on GMS contract.

No difference in immunisation coverage between practices (-1.08%, CI -17.95-15.8%)

Moderate (natural experiment – intervention dose unclear; likely difference between intervention and control practices; unclear if large enough sample; assessors not blinded; group effects not considered)

Norbury, 2011 [8]

315 general practices in Scotland

Eco B&A

Influenza in >65-year olds and risk groups

Incentive

QOF

Coverage in >65 yo and risk groups before (03-04) vs. after (06-07) introduction on QOF incentive in 2004.

Increase in coverage by 3.5% (CI 3.3 to 3.7%); higher in <65 yo 8.8% (CI 8.3 to 9.4%) than >65 yo 3.3% (CI 3.1 to 3.6%). Higher increases in those with disease risk, than age alone.

Low (no blinding of assessors)

Kontopantelis, 2012 [68]

All practices in England.

Eco CB&A

Influenza in people with CHD

Incentive

QOF

Coverage before vs. after the increase in upper payment threshold from 85% to 90% in 2006; and vs. other risk groups with no threshold change.

Immediate increase of 0.41% (CI 0.25 – 0.56%) population coverage, with larger increase seen in practices with <85% in 2006 of 0.85% (0.62 – 10.08%)

Low (no blinding of assessors)

Kontopantelis, 2014 [69]

Patients at 50 representative practices from 644 in CPRD

Eco ITS

Influenza in people with asthma

Incentive

QOF

Coverage before vs. after QOF target removed in 2006.

Small drop in coverage -0.70% (CI -1.1% to -0.39%

Low (no blinding of assessors)