Early impact of the COVID-19 pandemic and social distancing measures on routine childhood vaccinations in England, January to April 2020

Electronic health records were used to assess the early impact of COVID-19 on routine childhood vaccination in England to 26 April 2020. MMR vaccination counts fell from February 2020, and in the three weeks after introduction of social distancing measures were 19.8% lower (95% CI -20.7 to -18.9%) than the same period in 2019, before improving in mid-April. A gradual decline in hexavalent vaccination counts throughout 2020 was not accentuated on introduction of social distancing.


2
Electronic patient records from primary care were analysed to investigate changes in delivery of the routine childhood vaccination programme in England during the COVID-19 outbreak to 26 April 2020.

Near real-time records of routine childhood vaccinations from primary care
Two key milestones in the routine childhood immunisation programme delivered in primary care are first universal vaccinations at 8 weeks old, which include hexavalent vaccine (against diphtheria, tetanus, pertussis, polio, Haemophilus influenzae type b and hepatitis B), and vaccinations at one year old, which include the first dose of MMR vaccine (against measles, mumps and rubella). (1) Aggregated weekly counts of first hexavalent vaccinations delivered to infants under 6 months, and first MMR vaccinations delivered to children aged 12 to 18 months, were provided from TPP SystmOne for the first 17 weeks of 2019 and 2020. SystmOne is a software system which provides electronic patient records for over 2600 primary care practices in the UK and over 35 child health providers. Data were anonymous throughout, having been extracted as aggregated weekly counts grouped by local Clinical Commissioning Group for population level data checks.
To minimise changes in denominator, only providers active in SystmOne since 2018 contributed to the dataset. The majority of vaccinations entered into SystmOne are entered by general practices in real time (over 70% in 2019). However, vaccinations delivered in general practices which use other patient record software may be recorded at a delay into the SystmOne integrated patient record by local Child Health systems. To distinguish changes in vaccine delivery from lags in data recording, only vaccinations recorded on the same day as they were delivered were included.
The dataset included 69,568 hexavalent doses delivered in 2019, and 67,116 in 2020; and 68,849 MMR doses delivered in 2019, and 66,301 in 2020.

How did vaccination counts change during the COVID-19 outbreak?
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 11, 2020. . https://doi.org/10.1101/2020.05.07.20094557 doi: medRxiv preprint Hexavalent vaccination counts followed a similar pattern in 2020 to 2019, varying week-by-week, (particularly low counts in week 1 are likely explained by the national bank holiday and school holidays) (Figure 1). MMR vaccination counts also followed a similar pattern in 2020 until week 11, when they fell sharply and remained low for several weeks before rising again in weeks 16 and 17. The percent change in vaccinations counts in 2020 compared to 2019 varied over the course of the COVID-19 outbreak ( Table 1). Social distancing was a topic of public discussion in England from at least early March. (2) Prior to this, in weeks 1 to 9, hexavalent vaccination was 5.8% lower (95% CI -6.0 to -5.5%) and MMR vaccination 1.0% lower (95% CI -1.1 to -0.9%) in 2020 compared to 2019. On 12 March the UK government advised that anyone with a new continuous cough or a fever should self-isolate for seven days. In weeks 10 to 12 of 2020, hexavalent vaccination was 4.4% lower (95% CI -4.8 to -4.0%) and MMR vaccination 7.2% lower (95% CI -7.7 to -6.7%) than in 2019. Full social distancing measures were introduced on 20 March.(3) In the subsequent three weeks (weeks 13 to 15), hexavalent vaccination was 6.7% lower (95% CI -7.1 to -6.2%) and MMR vaccination 19.8% lower (95% CI -20.7 to -18.9%) than in 2019. In contrast, vaccination counts in 2020 were higher in weeks 16 and 17 of 2020 than for the same weeks in 2019, for both vaccines.
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2020. . https://doi.org/10.1101/2020.05.07.20094557 doi: medRxiv preprint    is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2020. . https://doi.org/10.1101/2020.05.07.20094557 doi: medRxiv preprint 6 with national holidays). However, a percent increase was also observed in week 17. This is promising for an early recovery in vaccination delivery, but will need monitoring to ensure it is sustained.
The greatest percent decrease in MMR vaccination was seen in London, which had a high early burden of COVID-19. However, decreases in MMR vaccination were also seen in regions with a low incidence of COVID-19 infection during this period, suggesting that the changes were not solely due to COVID-19 infection burden. SystmOne use varies regionally but it is unlikely that software system choice would result in vaccination counts being unrepresentative within regions.(6) Regional variation might also be explained by local communications, which may have mitigated the fall in some areas.
A limitation of this analysis is that changing counts of vaccinations delivered could be driven by numbers of eligible infants, rather than vaccine coverage. Decreasing birth rates may plausibly explain the gradually falling hexavalent vaccination counts,(7) and migration could play a role too, but these cannot explain the size and timing of the changes in MMR vaccination. Deferral of data entry could explain some decrease in real-time vaccination counts, but not the subsequent increase.
It is vital that routine childhood vaccinations continue, particularly for diseases such as measles for which a high coverage is required to prevent outbreaks. (5,8) Decreased vaccine counts have also been reported in other high-income countries,(9, 10) and the Regional Office for Europe of the World Health Organization has advised that routine immunization services should continue to aim for high population immunity.(11) Countries will require immunisation recovery plans with innovative approaches to delivery that maintain social distancing requirements. Vaccinations usually delivered in schools and affected by school closures will also require catch-up programmes.
Continuous and timely assessment of vaccine coverage will be required to respond to potentially volatile changes during the COVID-19 pandemic.
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2020. . https://doi.org/10.1101/2020.05.07.20094557 doi: medRxiv preprint Ethics approval: This analysis was conducted as part of public health usual practice, and was not conducted for research. Ethics approval was therefore not sought.

Conflict of interest statement
All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf  . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 11, 2020. . https://doi.org/10.1101/2020.05.07.20094557 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 11, 2020. . https://doi.org/10.1101/2020.05.07.20094557 doi: medRxiv preprint