Coronavirus as Recorded in Primary Care - EXPERIMENTAL STATISTICS

For comparison of different demographic groups, crude rates are shown as counts per 100,000. These are calculated over the time period and daily using, 

\({n_i \over N_i} \times{} 100000 \)

where \(n_i\) is the number of cases in group \(i\), and \(N_i\)is the total population of group \(i\). These denominators were calculated using a snapshot of the full GDPPR data set on 19th October 2020. Only patients with a valid LSOA, sex, and age, who were alive and registered at an English General Practice on that date were counted. The selected date is approximately midway through the study period and is between the two waves. The denominators were calculated for the demographic categories described in Section 4.  Ethnic category was assigned to the denominator using a methodology utilising both primary and secondary care data. Details on the methodology used assigning Ethnic group in the denominators are available here.

Figure 3: Total number of unique patients with a diagnosed or suspected case, as reported in primary care data, compared to the official cases by publication date, in England between 1st March 2020 and 31st March 2021.

Figure 3 presents the number of unique patients with a test or clinical diagnosis or suspected cases as compared to the official counts as presented on coronavirus.data.gov.uk. These graphs demonstrate a consistency in trends between the data sets. We can identify additional cases recorded in primary care during the earlier months of the pandemic when testing was not widely available. It should be noted that lateral flow tests are included in the official published data but are not routinely included in primary care records.

Figures  4 and 5 display the overlap between patients identified in primary care, pathways and SGSS test data. The two periods shown were selected to capture the change in SNOMED codes used to define infections, and to coincide broadly with the two waves of the virus as experienced in England. These diagrams demonstrate that no single data set captures all instances of confirmed or suspected COVID-19. During the initial months of the pandemic the majority of infections are identified through GP records whereas in the second half most are captured through testing.

Overall, during the report period, 5,670 per 100,000 (1 in 18) patients registered at a general practice, have a covid-19 positive test or diagnosis recorded. Including patients with suspected covid-19 diagnosis recorded increases this rate to 8,890 per 100,000 (1 in 11). Female patients have a higher rate of clinical and test Covid-19 diagnosis recorded, 6,325 per 100,000 (1 in 16) than male patients with 5,008 per 100,000 (1 in 20).

Figure 6: Unique patients with a clinical covid-19 diagnosis or test result by sex, per 100,000 population, as reported in primary care data, in England between 1st March 2020 and 31st March 2021.

Figure 7: Unique patients with a suspected or clinical covid-19 diagnosis or test result by sex, per 100,000 population, as reported in primary care data,  in England between 1st March 2020 and 31st March 2021

Figure 6 shows trends in unique patients with a clinical COVID-19 diagnosis or test result by sex as captured in general practice in England between 1st March 2020 and 31st March 2021. These are reported as 7 day rolling averages, per 100,000 of the underlying population.  Figure 7 additionally includes patients with a suspected diagnosis. Overall, there is an increase of cases during March that slowly decrease from mid-April until the end of June 2020. Rates gradually increase again between September and mid-November 2020.  Following a drop, we then observe a sharp rise from December 2020, with a sharp decrease from mid-January 2021.  We observe a consistently higher prevalence in the female population. These differences will reflect either differences in infection rates or differences in how the different groups tested or approached their GP.

Table 5 shows infection rates vary across ethnic groups between 1 March 2020 and 31 March 2021.  Patients in the Bangladeshi group have the highest rate of clinical diagnoses or positive tests recorded 9,608 per 100,000 (1 in 10). Patients in the Chinese group have the lowest  rate of clinical diagnoses and positive test recorded, 1,723 per 100,000 (1 in 58).

Figure 8: Unique patients with a clinical covid-19 diagnosis or test result by ethnic group, as reported in primary care data,  in England between 1st March 2020 and 31st March 2021.

Figure 9: Unique patients with a suspected, or clinical covid-19 diagnosis or test result by ethnic group, per 100,000 population, as reported in primary care data,  in England between 1st March 2020 to 31st March 2021.

Figure 8 shows trends in unique patients with a clinical COVID-19 diagnosis or test result by ethnic group as captured using an algorithm combining information collected in both general practice and secondary care, in England between 1st March 2020 and 31st March 2021. These are reported as 7 day rolling averages, per 100,000 of the underlying population. Patients are counted on their first clinical diagnosis or positive test result. Figure 9 additionally includes patients with a suspected diagnosis. In March and April 2020 overall rates increase with the Black Caribbean group being the highest and the Chinese group the lowest. Rates gradually increase again between September and mid-November 2020. Following a drop, a sharp rise is observed from December 2020, with a sharp decrease from mid-January 2021. Between September and December 2020, the Pakistani ethnic group have the highest rate whereas between December 2020 and February 2021 the highest is with the Bangladeshi ethnic group. These differences will reflect either differences in infection rates or differences in how the different groups tested or approached their GP.

Table 6 shows infection rates vary across geographic regions between 1 March 2020 and 31 March 2021.  Patients in the North West have the highest rate of clinical diagnosis or positive tests recorded 6,591 per 100,000 (1 in 15).  Patients in the South West have the lowest 4,113 per 100,000 (1 in 24).

Figure 10: Unique patients with a clinical covid-19 diagnosis or test result by geographic region, per 100,000 population, as reported in primary care data, in England between 1st March 2020 and 31st March 2021.

Figure 11: Unique patients with a suspected or clinical covid-19 diagnosis or test result by geographic region, per 100,000 population, as reported in primary care data, in England between 1st March 2020 and 31st March 2021

Figure 10 shows trends in unique patients with a clinical COVID-19 diagnosis or test result by geographic region as captured in general practice in England between 1 March 2020 and 31 March 2021. These are reported as 7 day rolling averages, per 100,000 of the underlying population. Patients are counted on their first clinical diagnosis or positive test result. Figure 11 additionally includes patients with a suspected diagnosis. In March and April 2020 overall rates increase with the East of England being the highest. Rates gradually increase again between September and mid-November 2020. Following a drop, a sharp rise is observed from December 2020, with a sharp decrease from mid-January 2021. Between September and December 2020, the North West, Yorkshire, and the North East have the highest rates whereas between December 2020 and January 2021 the highest is in London. These differences will reflect either differences in infection rates or differences in how patients in the different regions were tested or approached their GP. Figure 10 shows trends in unique patients with a clinical COVID-19 diagnosis or test result by geographic region as captured in general practice in England between 1 March 2020 and 31 March 2021. These are reported as 7 day rolling averages, per 100,000 of the underlying population. Patients are counted on their first clinical diagnosis or positive test result. Figure 11 additionally includes patients with a suspected diagnosis. In March and April 2020 overall rates increase with the East of England being the highest. Rates gradually increase again between September and mid-November 2020. Following a drop, a sharp rise is observed from December 2020, with a sharp decrease from mid-January 2021. Between September and December 2020, the North West, Yorkshire, and the North East have the highest rates whereas between December 2020 and January 2021 the highest is in London. These differences will reflect either differences in infection rates or differences in how patients in the different regions were tested or approached their GP.

Table 7 shows infection rates vary across deprivation quintiles between 1 March 2020 and 31 March 2021. These are assigned using IMD quintiles as described in Section  4. Patients in the most deprived quintile have the highest rate of clinical diagnoses or positive tests recorded 6,548 per 100,000 (1 in 15). Patients least deprive quintile have the lowest rate of clinical diagnoses and positive test recorded, 4,859 per 100,000 (1 in 21).

Figure 12: Unique patients with covid-19 clinical diagnosis or test result by IMD quintile, per 100,000 population, as reported in primary care data, in England between 1st March 2020 and 31st March 2021

Figure 13: Unique patients with a suspected or clinical covid-19 diagnosis or test result by IMD quintile, per 100,000 population, as reported in primary care data, in England between 1st March 2020 and 31st March 2021

Figure 12 shows trends in unique patients with a clinical COVID-19 diagnosis or test result by deprivation quintiles. These are reported as 7 day rolling averages, per 100,000 of the underlying population. Patients are counted on their first clinical diagnosis or positive test result. Figure 13 additionally includes patients with a suspected diagnosis.  Between the months of September and December 2020 the most deprived group has a higher rate compared to the other groups. These differences will reflect either differences in infection rates or differences in how the different groups tested or approached their GP.

Table 8 shows infection rates vary across age categories between 1 March 2020 and 31 March 2021. Patients in the over 90s group have the highest rate of clinical diagnoses or positive tests recorded 11,979 per 100,000 (1 in 8). Patients in the 5 to 11 group have the lowest  rate of clinical diagnoses and positive test recorded, 2,483 per 100,000 (1 in 40).

Figure 14: Unique patients with a covid-19 clinical diagnosis or test result by age category focusing on under 70s, per 100,000 population, in England between 1st March 2020 and 31st March 2021

Figure 15: Unique patients with a covid-19 clinical diagnosis or test result by age category focusing on over 70s, per 100,000 population, in England between 1st March 2020 and 31st March 2021

Figure 16: Unique patients with a suspected or clinical covid-19 diagnosis or test result by age category focusing on under 70s, per 100,000 population, in England between 1st March 2020 and 31st March 2021

Figure 17: Unique patients with a suspected or clinical covid-19 diagnosis or test result by age category focusing on over 70s, per 100,000 population, in England between 1st March 2020 and 31st March 2021

Figure 14 and Figure 15 show trends in unique patients with a clinical COVID-19 diagnosis or test result by age categories as captured in general practice on diagnosis date, in England between 1st March 2020 and 31st March 2021. These are reported as 7 day rolling averages, per 100,000 of the underlying population. Patients are counted on their first clinical diagnosis or positive test result. Figure 16 and Figure 17 additionally include patients with a suspected diagnosis.

Cases initially peak in April 2020, with the rate for the over-90s group being the highest. Rates for those aged 70+ remain higher than for other age bands from early in the pandemic through to summer 2020.

From September 2020 we observe differing trends across the age groups. There is a steep rise, then fall, for the 5-11 and, to a lesser extent, the 12-17 age groups in September 2020. This is driven by suspected cases rather than clinical diagnoses or test results.  In early October the 18-25 age group saw a steep rise then remained relatively flat through mid-November 2020, while other age groups increased more steadily, and to lower peaks, over this period.

From mid-November through early-December 2020 all age bands saw some declines, then rising again to peaks in mid-January 2021. The pattern across age bands differs from those seen earlier in the pandemic. For bands covering the ages 12 to 69, January 2021 saw rates markedly higher than those seen in April 2020. For the 70+ groups, rates were lower in January 2021 than in April 2020. Unlike in the earlier period, peak rates for those aged 70-79 and 80-89 were lower than for younger adult age bands, although the rate for the over-90s group was again the highest across all age bands. After January 2021 there was a sharp decline for all age groups.

These differences will reflect either differences in infection rates or differences in how the different groups tested or approached their GP.