A cross-sectional survey of the 135 CCGs in England was conducted to identify how face-to-face general practice consultations were being delivered across the country.

Survey design

Survey questions were devised by the study team. They concerned models of face-to-face consultations in use and the patient populations each were available to; prior use of the hub model; and planned evaluations. Questions were pre-tested with a researcher experienced in survey design, two CCGs and one provider of primary healthcare. Minor changes to wording were made for clarity. The final questionnaire is available as Extended data. ¹³

Data collection

Questions were sent by email to all CCGs in June 2020 under the Freedom of Information (FOI) Act 2000. This legislation enables public access to recorded information held by public authorities in England. Responsibility for cleansing data lies with the authorities responding to FOI requests ⁵ and research ethics approval was not required.

Individual CCGs were identified on the NHS England website and their specific FOI procedures were followed. FOI regulations mandate a response timeframe of 20 working days. Where, rarely, replies were not received within this time, follow-up emails were sent, and telephone calls made if necessary. All responses were collated in an Excel spreadsheet for analysis.

Data analysis

Data was extracted for each survey question, with any queries regarding response interpretation discussed between the authors. Email and telephone communication with CCGs was also used occasionally for clarification or updates. Numeric and narrative analyses were undertaken for responses to each question. Different face-to-face delivery models were identified, and each CCG was assigned to a model combination of best fit for nationwide comparison.

Specific data on clusters or outbreaks of acute respiratory infections with at least one linked case testing positive for SARS-CoV-2 are recorded for hospitals and care homes in the weekly COVID-19 surveillance reports , published by Public Health England (PHE). ^{6, 7} [‘Clusters’ are cases linked to a setting through surveillance testing; ‘outbreaks’ are reported directly from settings]. To obtain similar information regarding general practice, an email was sent to PHE in December 2020, under the FOI Act 2000. The questions asked are available as Extended data. ¹⁴

Responses

Replies were received from all CCGs, 99% in June and July 2020, with the final response received on 2 ^nd October.

Response interpretation

Terminology used in responses varied – ‘hot sites’, ‘hot clinics’ and ‘resilience hubs’ could be used to refer to the same service for example. Colour-coding, e.g., ‘green’ and ‘amber’ sites, could also be used differently, and service provision for non-COVID patients was not always clear. As a consequence the distinction between hot hubs, cold hubs and zoned practices was not always immediately evident. Complete response sets, including any additional documentation provided on model pathways and usage data, were therefore used to interpret and categorise all face-to-face consultation types according to the models in this paper.

The following interpretation of the data is the authors’ own and has not been approved or otherwise by the CCGs.

Adapted delivery models

General practice face-to-face delivery has been modified in each CCG using combinations of the three models previously described: (i)

Zoned practices (model 1 in Figure 1), available to the entire patient populations served, were reported by 63 (47%) of CCGs. All but 12 of these also used hubs at the time of reporting. Most commonly, two closed ‘red’ and ‘green’ areas with different entrances and exits were used. Rarely, cohorts could also be separated temporally, with COVID-19 patients alone seen at certain times. The latter model was described in updated NHS England guidance (version 2, dated 5 ^th April 2020) for surgeries where provision of separate spaces for different cohorts was not possible. ²

The different means of delivering face-to-face services were compared for each CCG. Overall approaches taken were found to fit one of seven model combinations, albeit with some distinctions, notably the different usage of home visits and co-located hubs.

The seven model combinations are illustrated in Figure 1, and their distribution among CCGs shown in Figure 2. The diverse spread of models and size of CCGs seen in Figure 2 reflect the varied geography and population across England. A degree of cross-CCG working was evident in replies, and a small number of CCGs had access to hubs across neighbouring boundaries. Figure 1.

Authors’ interpretations of CCG responses according to the following definitions: •

Zoned practice: co-location of hot and cold services on a single site, serving own practice list

PHE data received detailed regional outbreaks or clusters of acute respiratory infections with possible/confirmed COVID-19 cases linked to GP surgeries. Table 1 shows this data as monthly totals from March to December 2020, alongside numbers of English face-to-face GP consultations obtained from NHS Digital. ⁸ In the March-August timeframe covered by model usage in this study, 25 outbreaks or clusters with possible/confirmed COVID-19 cases were reported in England in the context of over 104 million face-to-face GP appointments. This represented less than 2% of all COVID-19 outbreaks or clusters in England [calculated using data in COVID-19 surveillance reports]. ^{6, 7} Numbers increased in October and November, alongside rising face-to-face appointment levels, but continued to represent a small proportion of totals. ^{6, 7} Final figures may in fact be lower after investigation of unconfirmed cases by local PHE Health Protection Teams. Table 1.

Month *	North East	Yorkshire and Humber	North West	East of England	East Midlands	West Midlands	South East	South West	London	England	Monthly GP face-to-face appointments, England **
March	0	0	0	0	0	0	0	0	0	0	31,841,558
April	0	0	0	0	0	0	1	1	0	2	14,960,202
May	0	0	1	5	0	0	0	0	1	7	15,459,694
June	0	3	4	0	2	0	2	0	2	13	19,526,124
July	0	1	0	0	0	1	0	0	1	3	22,352,108
Aug	0	2	2	0	0	0	0	2	1	7	20,763,100
Sept	3	1	2	0	2	2	1	0	3	14	30,185,304
Oct	1	9	25	5	8	23	7	18	3	99	33,741,240
Nov	10	8	9	10	12	24	13	3	13	102	27,773,964
Total	14	24	43	20	24	50	24	24	24	247	216,603,294

*

Monthly data calculated from weekly totals given in Public Health England response to FOI request. [March: 02/03/2020 to 29/03/2020; April: 30/03/2020 to 26/04/2020; May: 27/04/2020 to 31/05/2020; June: 01/06/2020 to 28/06/2020; July: 29/06/2020 to 26/07/2020; Aug: 27/07/2020 to 30/08/2020; Sept: 31/08/2020 to 27/09/2020; Oct: 28/09/2020 to 01/11/2020; Nov: 02/11/2020 to 29/11/2020]. Some 'possible' COVID-19 cases reported may be excluded after investigation by local PHE Health Protection Teams, and final numbers may therefore be lower.

All CCGs reported the use of zoned practices, hubs and/or home visits in varying combinations for face-to-face GP consultations. Factors influencing model selection included COVID-19 appointment demand, presence of pre-established collaborative networks and GP preferences to provide continuity of care. On-going model assessments enabled responsiveness to changing demand, mainly through altered hub availability, but also by changed provision of home monitoring for COVID-19 patients.

CCG mergers on 1st April 2020, which decreased total numbers from 191 to 135, may also have impacted model patterning within CCGs. While the ‘hot hubs + cold practices’ model combination #2 was the most frequently used nationwide, larger CCGs tended to report ‘mixed’ models, giving model combination #5 the greatest overall coverage ( Figure 2). Patterning within these mixed model combinations (#5-#7) could not always be identified, but in one CCG five areas of zoning and a sixth using a hot hub were identified, and in some others more diffuse patterning was noted.

The distinction between the zoned practice and hub models used is not as clear as indicated in our Introduction. The co-location of hot hubs with cold services means that the requirement for strict management between hot and cold areas is as important here as in the zoned practice model. Indeed, several CCGs using co-located hubs and some using zoned practices specified the use of separate entrances and exits, with some also indicating separate car parking facilities. By contrast, other zoned practices shared more similarities with distinctly sited hubs, where red and green areas were split between main and branch surgeries for example, or additional structures such as portacabins were used on site, to separate patient cohorts registered within one practice. Thus, it is not necessarily the case that the hub model provides clearer separation and thereby simpler IPC adherence than zoned practices, as indicated in the guidelines. ^{2, 4}

Usage of home visits also varied. In at least two CCGs this was the main or only form of COVID-19 face-to-face consultation. In 30% of CCGs however, the use of specific hot and/or cold home visits was not reported. Home monitoring, via pulse oximetry for example, was an additional service indicated by a small number of CCGs at the time of reporting.

While the risk of transmission can be minimised in face-to-face consultations, it cannot be eliminated, particularly with the significant pre-symptomatic and asymptomatic transmission associated with COVID-19. ⁹ PHE data shows a total of 247 possible or confirmed COVID-19 outbreaks or clusters in English GP surgeries over the first nine months of the pandemic, in the context of more than 216 million face-to-face appointments. While these numbers are a small proportion of total (all source) outbreaks in the UK, investigation of any confirmed outbreak may guide future measures to reduce transmission.

The variability in possible/confirmed outbreaks or clusters in general practice over time, as well as by region (shown in Table 1), may partly be accounted for by changing COVID-19 incidence rates, but also by other factors including differences in testing rates, population risk factors, and the numbers and means of delivery of face-to-face consultations taking place locally.

In addition to minimising SARS-CoV-2 transmission, the success of IPC delivery models used during the pandemic will be measured in terms of the management of COVID-19 and other conditions, and the perspectives of patients and staff. Several patient surveys have now been undertaken. ^{10 – 12} The authors have also reported on a survey of the public, investigating experiences and understanding of the changes to general practice. ¹⁵ These patient findings, taken together with staff perspectives ³ and quantitative evaluations should provide helpful insights into the effective use of the adapted models.

There are several limitations to this study. As already described, terms used to describe novel concepts (e.g., hot/cold hubs, zoning) varied and could not readily be checked. This may, for a small number of CCGs, have led to misinterpretation of responses and consequent misassignment to model combinations. Studies at Primary Care Network or practice level would likely reveal more granular detail, particularly with respect to the ‘mixed model’ combinations identified. In terms of the PHE FOI request, data on outbreaks/clusters from ‘GP surgeries’ was provided, and it is unclear whether this includes hubs and all home visiting services. Finally, the service delivery details provided by CCGs were mainly received in summer 2020 and, as has been identified, these are dynamic models and may since have been modified.