{"paper_id":"021474ac-48cf-4972-8548-86fd6f08ce30","body_text":"OpenSAFELY NHS Service Restoration Observatory 2: changes in \nprimary care activity across six clinical areas during the COVID-19 \npandemic \n \nHelen J Curtis1*, Brian MacKenna1*, Milan Wiedemann1, Louis Fisher1, Richard Croker1*, \nCaroline E Morton1, Peter Inglesby1, Alex J Walker1, Jessica Morley1, Amir Mehrkar1, \nSebastian CJ Bacon1, George Hickman1, David Evans1, Tom Ward1, Simon Davy1, William J \nHulme1, Orla Macdonald1, Robin Conibere2, Tom Lewis3, Martin Myers4, Shamila \nWanninayake5, Kiren Collison6, Charles Drury7, Miriam Samuel8, Harpreet Sood9, Andrea \nCipriani10,11, Seena Fazel10,11, Manuj Sharma12, Wasim Baqir6, Chris Bates13, John Parry13, \nBen Goldacre1  \n \n \n(*These authors contributed equally to this work) \nCorresponding: ben.goldacre@phc.ox.ac.uk  \n \n \n1 The Bennet Institute for Applied Data Science, Nuffield Department of Primary Care Health \nSciences, University of Oxford, OX2 6GG, UK \n2 Beacon Medical Group, Plymouth, UK \n3 Northern Devon Healthcare NHS Trust, UK \n4 Lancashire Teaching Hospitals NHS Foundation Trust, UK \n5 The Manor Surgery, Oxford, UK \n6 NHS England and NHS Improvement, London, UK \n7 Herefordshire and Worcestershire Health and Care NHS Trust, UK \n8 Wolfson Institute of Population Health (Queen Mary University of London) \n9 University College London Hospitals NHS Foundation Trust \n10 Department of Psychiatry, University of Oxford, Oxford \n11 Oxford Health NHS Foundation Trust, Warneford Hospital, Oxford \n12 Department of Primary Care and Population Health, University College London. \n13 TPP, TPP House, 129 Low Lane, Horsforth, Leeds, LS18 5PX \n \nAbstract \n \nBackground \nThe COVID-19 pandemic has disrupted healthcare activity across a broad range of clinical \nservices. The NHS stopped non-urgent work in March 2020, later recommending services be \nrestored to near-normal levels before winter where possible.  \n \nAims \nUsing routinely collected data, our aim was to describe changes in the volume and variation \nof coded clinical activity in general practice in: (i) cardiovascular disease, (ii) diabetes, (iii) \nmental health, (iv) female and reproductive health, (v) screening, and (vi) processes related \nto medication. \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \nNOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.\n\n \n1 \nDesign and setting \nWith the approval of NHS England, we conducted a cohort study of 23.8 million patient \nrecords in general practice, in-situ using OpenSAFELY. \n \nMethods \nWe selected common primary care activity using CTV3 codes and keyword searches from \nJanuary 2019 - December 2020, presenting median and deciles of code usage across \npractices per month. \n \nResults \nWe identified substantial and widespread changes in clinical activity in primary care since \nthe onset of the COVID-19 pandemic, with generally good recovery by December 2020. A \nfew exceptions showed poor recovery and warrant further investigation, such as mental \nhealth, e.g. “Depression interim review” (median across practices in December 2020 -41.6% \ncompared to December 2019). \n \nConclusions \nGranular NHS GP data at population-scale can be used to monitor disruptions to healthcare \nservices and guide the development of mitigation strategies. The authors are now \ndeveloping real-time monitoring dashboards for key measures identified here as well as \nfurther studies, using primary care data to monitor and mitigate the indirect health impacts of \nCovid-19 on the NHS. \n \n \n \nHow this fits in \nDuring the COVID-19 pandemic, routine healthcare services in England faced significant \ndisruption, and NHS England recommended restoring NHS services to near-normal levels \nbefore winter 2020. Our previous report covered the disruption and recovery in pathology \ntests and respiratory activity: here we describe an additional six areas of common primary \ncare activity. We found most activities exhibited significant reductions during pandemic wave \n1 (with most recovering to near-normal levels by December); however many important \naspects of care - especially those of a more time-critical nature - were maintained throughout \nthe pandemic. We recommend key measures for ongoing monitoring and further \ninvestigation of the impacts on health inequalities, to help measure and mitigate the ongoing \nindirect health impacts of COVID-19 on the NHS.  \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n2 \nBackground \nThe COVID-19 pandemic disrupted healthcare services globally.1 In March 2020, NHS \nEngland initially promoted measures to reduce viral transmission and provide only essential \nhealth services.2,3 The World Health Organisation (WHO) recommended rapid assessments \nof healthcare capacity and the development of key performance indicators.4,5 From August \n2020, NHS England aimed to restore primary care and other services back to normal activity \nwhere clinically appropriate.6  \n \nVarious studies have assessed the impact of the pandemic on non-COVID health services, \nfor example mental health, female and reproductive health, health screenings, and \nprescribing.\n7–10 Overall, findings suggest a substantial decrease from the first official \nlockdown in the UK (March 2020), with recovery to near-normal activity in most clinical areas \nfrom around July 2020. We previously reported a data-driven approach for monitoring \nhealthcare disruptions and recovery using the OpenSAFELY platform \n(https://www.opensafely.org/\n) combined with input from a clinical advisory group, intended to \nexplore changes in high volume areas, including those that might otherwise be missed.11 \nSelecting pathology tests and respiratory conditions as key examples, we showed that \nactivity largely decreased substantially and subsequently recovered. However, we also \nfound that some activities such as blood coagulation tests were well-maintained, suggesting \nthat important clinical care was effectively prioritised.11 Certain clinical conditions like \ncardiovascular disease (CVD) and diabetes are associated with higher risk of morbidity and \nmortality from COVID-19, emphasising the importance of maintaining good routine care.12–15  \n \nWe therefore set out to generate an overall picture of changes in clinical activity in primary \ncare across key areas of medicine; and to identify key measures to continuously monitor the \nimpacts of COVID-19 on the NHS and inform further studies. Specifically we extended our \nearlier study\n11 to December 2020 and expanded our work to cover six further clinical areas: \ncardiovascular disease, diabetes, mental health, female and reproductive health, screening \nprocedures, and processes related to medication.   \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n3 \nMethods \nStudy design and data source \nFollowing the methodology previously described11, we conducted a cohort study using \nroutinely collected pseudonymised primary care electronic health records (EHR) in the \nOpenSAFELY-TPP platform (https://opensafely.org). This included 40% of England’s \ngeneral practices (2,500, those using TPP SystmOne software), covering 24m patients. In \nbrief, this structured GP data includes one record for every diagnostic code, prescription, \nblood test, investigation, or similar in primary care.  \nStudy population and data processing \nThe study population and data processing were as in our previous study11, except for a later \nstudy end date (31st December 2020), and included all patients registered with a general \npractice on 31st December 2020. Briefly, we counted occurrences of all coded clinical \nactivities per month throughout 2019-2020, by Clinical Terms Version 3 (CTV3) code \n(including diagnoses, investigations and other clinical/administrative processes, but \nexcluding medications/vaccinations), grouped by general practice. We excluded any codes \nwith less than 1000 total occurrences in 2020 to identify the most frequent clinical activities. \nWe used the CTV3 parent-child hierarchy (e.g. 24: ‘Examination of cardiovascular system’; \n242: ‘O/E - pulse rate’) to group similar activities, and mapped each code to high level CTV3 \nconcepts to assist with categorisation into broad topics (e.g. “cardiovascular”). For each \ncode/group we calculated the monthly rate of code usage per 1000 registered patients and \nthe median and deciles across practices.  \nStudy measures \nWe pragmatically grouped activity and selected clinical codes relevant to each of the \nfollowing topics: Cardiovascular disease, Diabetes, Mental health, Female and reproductive \nhealth, Screening and related procedures, and Processes related to medication. We did not \ninclude prescribing, which is coded in GP systems using the NHS dictionary of medicines \nand devices, as high quality routinely updated analysis of primary care dispensing data is \nalready openly available on our partner service OpenPrescribing \n(https://openprescribing.net/\n). The selection of clinical codes was largely based on existing \nCTV3 concepts and keyword searching. A detailed description of our methodology is \navailable in Supplementary Information and Table S1.  \nClinical advisory group  \nWe established a clinical advisory group to review our findings, consisting of general \npractitioners, pharmacists, pathologists, other relevant specialists, and national clinical \nadvisors, formed by invitation of clinicians known through existing professional relationships. \nThe raw results of our data driven approach on each topic were discussed with the advisory \ngroup during a series of online meetings to prioritise clinical topics and inform interpretation. \nThe group also had the opportunity to comment on these documents outside of the \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n4 \nmeetings, and if needed arrange further meetings with the research team. Additionally the \ngroup was asked to select “key measures” of activity from each clinical area to direct \ntargeted work on health inequalities. \nSoftware and reproducibility \nData management and analysis were performed using Python 3.8. Code for data \nmanagement and analysis are available online (https://github.com/opensafely/restoration-\nobservatory-data-driven). \nPatient and public involvement \nWe have developed a publicly available website https://opensafely.org/ through which we \ninvite any patient or member of the public to contact us regarding this study or the broader \nOpenSAFELY project. \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n5 \nResults \nAll included clinical codes/groups for each topic along with their total 2020 usage are shown \nin Tables S2-S7. We summarise results by topic, highlighting selected code usage at key \npoints in the pandemic in England (February, April and December 2020; Table 1).  \n \nTable 1. Rate of recording of selected CTV3 codes and groups of codes for each topic in \nFebruary, April and December 2020. (a) cardiovascular disease (CVD), (b) diabetes, (c) mental \nhealth, (d) female & reproductive health, (e) screening procedures, and (f) processes related to \nmedication. Showing median rate across English practices per 1000 registered patients and % \nchange on same month in previous year. Further detailed breakdowns of code recording are shown in \nTable S2.\n \nTopic CTV3 code/group \nMedian rate per 1000 registered patients \n(% change on previous year) \nFebruary April December \n(a) CVD \n“24” - Examination of cardiovascular \nsystem 164.2 19.8 \n(-86.9%) \n82.6 \n(-42.6%) \n“18” - Cardiovascular symptoms (& \n[heart]) 0.6 0.3 \n(-57.1%) \n0.4 \n(-17.7%) \n“G5” - Other forms of heart disease 0.5 0.3 \n(-48.9%) \n0.4 \n(-6.4%) \n“XaQVY” - QRISK2 cardiovascular \ndisease 10 year risk score 6.9 0.6 \n(-98.3%) \n3.0 \n(-38.9%) \n\"XaKFx\" - Average home systolic blood \npressure\n 0.4 0.2 \n(-8.1%) \n0.7 \n(+102.8%) \n\"XaKFw\" - Average home diastolic blood \npressure\n 0.4 0.2 \n(-8.3%) \n0.6 \n(+98.0%) \n(b) Diabetes \n\"66A\" - Diabetic monitoring 4.6 0.4 \n(-87.5%) \n3.3 \n(-16.6%) \n\"XaPbt\" - Haemoglobin A1c level - IFCC \nstandardised\n 28.2 3.4 \n(-86.2%) \n22.4. \n(-0.6%) \n\"XaIIj\" - Diabetic retinopathy screening 1.8 0.0 \n(-100.0%) \n0.7 \n(-53.7%) \n\"XaIeH\" - O/E - Right diabetic foot at low \nrisk 3.8 0.0 \n(-100.0%) \n2.7 \n(-18.0%) \n\"XaIuE\" - Diabetic foot examination 0.2 0.0 \n(-100.0%) \n0.2 \n(+113.8%) \n(c) Mental \nHealth \n\"E2\" - Neurotic, personality and other \nnonpsychotic disorders 1.2 0.6 \n(-44.5%) \n0.8 \n(-23.7%) \n\"E20\" - Neurotic disorder 0.8 0.5 \n(-39.1%) \n0.5 \n(-21.9%) \n\"XaK6f\" - Depression interim review 0.9 0.4 \n(-56.2%) \n0.6 \n(-41.6%) \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n6 \n\"XE0re\" - Depressed mood 0.6 0.2 \n(-64.7%) \n0.4 \n(-19.9%) \n(d) Female & \nReproductive \nhealth \n\"61\" - Contraception 3.2 1.2 \n(-60.6%) \n3.0 \n(+12.1%) \n\"15\" - [Gynaecological] or [obstetric] \nhistory 1.7 0.8 \n(-51.3% ) \n1.3 \n(-17.2%) \n\"64\" - (Child health care)(inf feed \nmeth)(breast/oth feed diff age)\n 1.1 0.8 \n(-33.0%) \n1.0 \n(-11.8%) \n\"K3\" - Disorder of breast 0.2 0.1 \n(-57.9%) \n0.2 \n(+17.8%) \n\"621\" - Patient currently pregnant 0.9 0.7 \n(-26.6%) \n0.8 \n(-8.0%) \n(e) Screening \n\"XaPVj\" - Bowel cancer screening \nprogramme: faecal occult blood result 6.4 3.0 \n(-52.1%) \n6.5 \n(+12.0%) \n\"Xa8Pl\" - Cervical smear 4.8 0.2 \n(-95.6%) \n3.9 \n(+20.9%) \n\"XaRBQ\" - NHS Health Check completed 1.8 0.0 \n(-100.0%) \n0.0 \n(-100.0%) \n(f) Processes \nrelated to \nmedication \n\"XaF8d\" - Medication review done 22.7 14.6 \n(-31.0%) \n18.7 \n(-8.2%) \n\"XaJKW\" - Patient understands why \ntaking all medication\n 0.1 0.0 \n(-100.0%) \n0.4 \n(+422.2%) \n \nCardiovascular disease (CVD) \nThe majority of cardiovascular disease coded activity experienced a substantial decline \nduring the initial stages of the pandemic, with limited recovery by September 2020 that \nlevelled off through to December 2020, e.g. blood pressure recording and \nelectrocardiography (Figure 1a-b).  \n \nExceptions included symptoms related to cardiovascular system (Figure 1c), which \nexperienced a sustained drop (April -57.1%, December -17.7%); “other forms of heart \ndisease” (Figure 1d), including atrial fibrillation and heart failure, experienced a small drop \nbut largely recovered (April -48.9%, December -6.4%); QRISK2 (XaQVY), a CVD risk tool, \ndropped dramatically (-98.3%) with limited recovery (December -38.9%). Blood pressure at \nhome codes increased overall (+100% in December, Figure 1f), but were infrequently used \n(410k, 400k total events, Table S2) compared to blood pressure codes recorded with no \nsetting (11 million, Figure 1a). \n \n \n \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n \n a   24 “Examination of cardiovascular \nsystem     (& [vascular system])”        \n    \n b  32 “Electrocardiography” \n \n \n \n \nc  18 “Cardiovascular symptoms (& [heart])” \n \nd  G5 “Other forms of heart disease” \n  \ne  XaQVY “QRISK2 cardiovascular disease \n10 year risk score” \nf  XaKFx “Average home systolic blood pressure” \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n  \nFigure 1. Recording of grouped subsets of cardiovascular CTV3 codes across TPP practices in England (January \n2019 - December 2020). (a) “Examination of cardiovascular system (& [vascular system])”. (b) \n“Electrocardiography”, (c) “Cardiovascular symptoms (& [heart])”, (d) “Other forms of heart disease”, (e) “QRISK2 \ncardiovascular disease 10 year risk score”, (f) “Average home systolic blood pressure”. Each group includes \nCTV3 codes that begin with the digits shown and is not necessarily an exhaustive collection of every activity \nrelated to the description. For grouped codes (a-d), the top 5 codes represented within each group are listed in \ntables below each graph. \nDiabetes \nDiabetic monitoring and HbA1c testing both experienced a large drop with good recovery \n(April -87.5%, December -16.6%; April -86.2%, December -0.6%; Table 1b, Figure 2a-b). For \nDiabetic Retinopathy Screening, the median dropped to zero in April, with limited recovery \nby December (-53.7%; Figure 2c). For “right diabetic foot at low risk”, the median dropped to \nzero in April, with good recovery by December (-18.0%; Figure 2d). There was substantial \nvariation in the rate of diabetes monitoring and retinopathy screening at baseline, indicating \nsome incompleteness. \n \n a   66A - “Diabetic monitoring” b  XaPbt - “Haemoglobin A1c level - IFCC standardised” \n \nc   XaIIj - “Diabetic retinopathy screening” \n \n   d\n   XaIeH - “O/E - Right diabetic foot at low risk” \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n \nFigure 2. Recording of CTV3 codes across TPP practices in England (January 2019 - December 2020). (a)  \n“Diabetic monitoring”, (b) “Haemoglobin A1c level - IFCC standardised”, (c) “Diabetic retinopathy screening”, (d) \n“O/E - Right diabetic foot at low risk”. Each code is not necessarily an exhaustive collection of every activity \nrelated to the description. O/E = “on examination”. “O/E - Left diabetic foot at low risk” is not shown but has an \nalmost identical pattern to the right foot equivalent. Only ‘at low risk’ codes were included as codes related to \nmoderate or increased risk did not meet the frequency threshold for the data driven analysis. For grouped codes \n(a), the top 5 codes represented within the group are listed under the graph. \n  \nMental health \nThe majority of mental health coded activity experienced a moderate decline during the initial \nstages of the pandemic, with incomplete recovery by December 2020 (Table 1c), e.g. \n“Neurotic, personality and other nonpsychotic disorders”: April -44.5%, December -23.7% \n(Figure 3a); “Depressed mood”: April -64.7%, December -19.9% (Figure 3b). “Depression \ninterim review” activity showed particularly poor recovery (April -56.2%, December -41.6%, \nFigure 3c). \n \n \n \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n \n a   E2 - “Neurotic, personality and other \nnonpsychotic disorders” \nb  XE0re - “Depressed mood” \n \n \n \nc   XaK6f - “Depression interim review” \n \n    \n \n \nFigure 3. Recording of CTV3 codes across TPP practices in England (January 2019 - December 2020). (a)  \n“Neurotic, personality and other nonpsychotic disorders”, (b) “Depression interim review”, (c) “Depressed mood”. \nEach code is not necessarily an exhaustive collection of every activity related to the description. For grouped \ncodes (a), the top 5 codes represented within the group are listed under the graph.\n \nFemale and reproductive health \nThe majority of female and reproductive-related activity experienced a moderate decline \nduring the initial stages of the pandemic, with either recovery to near-normal levels by \nDecember 2020 or a return to an existing increasing or decreasing trend; for example \n“Contraception” (codes beginning “61”): April -60.6%, December +12.1% (Figure 4a).  \n \n“[Gynaecological] or [obstetric] history” showed a broadly similar pattern but with a slight \nsustained reduction (December -17.2%, Figure 4b). Various codes encompassed \ngynaecological and obstetric procedures and symptoms, which generally recovered to pre-\npandemic levels, although many had a median of zero throughout the period. One example \nwas “Disorder of breast” (Figure 4c, the most common code within which was “Breast lump”), \nwhich had a small increase by December (April -57.9%; December +17.8%).  \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n11 \nCodes related to infant health such as six-week checks only had a small drop and recovered \nto near-normal levels, perhaps following a slightly decreasing trend (Figure 4d).  \n \nCodes for “Patient currently pregnant” reduced slightly and a small reduction was maintained \nthrough to December (Figure 4e). \n \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n \n a  61 - “Contraception”  b  15 - “[Gynaecological] or [obstetric] history” \n  \nc K3 - “Disorder of breast” d 64 - “(Child health care)(inf feed meth)(breast/oth \nfeed diff age)” \n  \n \nc  621 - “Patient currently pregnant” \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n \n \n \nFigure 4. Recording of grouped subsets of female and reproductive health-related CTV3 codes across TPP \npractices in England (January 2019 - December 2020). (a) “Contraception”. (b) “[Gynaecological] or [obstetric] \nhistory”, (c) “Patient currently pregnant”. Each CTV3 code does not necessarily represent all activity related to the \ndescription. The top 5 codes represented within each group are listed under the graph.\n \nScreening and related activity \nThe majority of screening and related activity experienced a substantial decline, with \nrecovery to slightly above normal levels by December 2020 (Table 1e), e.g. “Bowel cancer \nscreening programme: faecal occult blood result” (April median -52.1%, December +12.0%) \nand “Cervical screening” (April -95.6%, December +20.9%). However, the patterns of \nrecovery differed, with bowel cancer screening only beginning to recover around September \nwhile cervical screening was near normal by July (Figure 5a-b).  \n \nNHS health checks reduced from 1.8 per thousand in February to close to zero activity in \nApril 2020 (median 0.0, with some recovery but median remaining zero by December 2020 \n(Figure 5c). Alcohol screening had a skewed distribution with a median of only 0.1 records \nper thousand in February 2020, reducing to zero in April and December 2020.  \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n a  XaPVj -  “Bowel cancer screening \nprogramme: faecal occult blood result” \n b  Xa8Pl -  “Cervical smear” \n \n \nc  XaRBQ - “NHS Health Check completed” \n \n \n \n \nFigure 5. Recording of screening-related CTV3 codes across TPP practices in England (January 2019 - \nDecember 2020). (a) “Bowel cancer screening programme: faecal occult blood result”. (b) “Cervical Screening”, \n(c) “NHS Health Check completed”. Each CTV3 code does not necessarily represent all activity related to the \ndescription. Note, Diabetic retinopathy screening was also included with screening codes, but is discussed in the \nDiabetes section. \nProcesses related to medication \nSeveral codes explicitly mentioned “medication review”, most commonly XaF8d “Medication \nreview done” (7.25m occurrences, Figure 6a), which experienced a small dip and gradual \nrecovery (April median -31.0%, December -8.2%). Other medication review codes (e.g. \nreview by pharmacist, for a specific disease, or indicating presence/absence of patient), \nwere used less consistently between practices.  \n \nSeveral codes, generally uncommon but increasing in usage, were likely recorded as part of \na medication review, such as “Patient understands why taking all medication” (Figure 6b; \nFebruary median 0.1, April 0.0, December 0.4), and “Able to manage medication” (Xa2yC). \n \nFrom September 2020 Structured medication reviews appeared (Figure 6c) and increased \nrapidly to around 350k records per month; however, some practices were recording them at \nmuch higher rates than others (Figure 6d). \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n a   XaF8d - “Medication review done”  b XaJKW - “Patient understands why taking all \nmedication” \n  \nc Total activity per month in Structured \nMedication Reviews \nd Practice activity rate per month in \nStructured Medication Reviews \n \n \nFigure 6. Recording of CTV3 codes across TPP practices in England. (a) “Medication review done” (January \n2019 - December 2020) (b) “Patient understands why taking all medication”.(c) Total recording of Structured \nMedication Reviews across TPP practices in England throughout their period of use to date (September 2020 - \nDecember 2020), and (d) practice deciles showing the rate per 1000 registered patients for Structured Medication \nReviews. Each code is not necessarily an exhaustive collection of every activity related to the description.   \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n16 \nDiscussion \nSummary \nWe identified widespread but also heterogeneous changes in clinical activity in primary care \nsince the onset of the COVID-19 pandemic. There was generally good recovery by \nDecember 2020, with some exceptions: notably mental health, which showed minimal \nrecovery. There was also variation from the median across practices, for both baseline and \nrecovery. \nStrengths and weaknesses \nThe key strengths are the scale and completeness of the underlying raw EHR data, available \nclose to real time, and we engaged with clinicians for added context. All processed data and \nanalytical code is openly available in the Supplementary Materials or Github. We will publish \nour recommended key measures in a live updating report, and we encourage other groups \nto use OpenSafely for further exploration. Our data-driven approach is intended to generate \nan overall picture of primary care clinical activity, and explore high volume areas that might \notherwise be missed, for example when not included in manually curated codelists. \n \nDespite the strengths we recognise some limitations as previously discussed.\n11 Our data-\ndriven approach and filtering processes may have omitted some relevant codes; codes do \nnot necessarily indicate unique or new events, and may be affected by changes in coding \nbehaviour. All coded activity for patients registered at the end of the study period were \nincluded, and all activity was included under their latest practice. Patients who died or \nderegistered from TPP practices during the study period were not included. Overall, activity \ncounts were up to 6-8% lower than database totals in the earliest months of the study period.  \n \nInterpretation and context for each clinical area \nGiven the diversity of clinical areas covered by this overarching analysis, the clinical advisory \ngroup evaluated and interpreted the variation for each clinical area separately. \nCVD \nMuch coded activity related to monitoring, and remained around 40% reduced from pre-\npandemic levels. This was not surprising due to changes in guidance and financial \nincentives.16 We warn that electrocardiogram data should be interpreted cautiously as this is \noften conducted outside primary care and not always systematically coded. The lack of \nrecovery in QRISK2 scores may have public health significance, potentially causing later \ndiagnosis of heart disease and poorer early management. Home blood pressure coding \nunsurprisingly increased, however home monitoring in general may not always be recorded \ncompletely or consistently in GP records. The consistent pattern of decrease in most \ncardiovascular related activity is in line with results from other studies in the UK.\n8,17,18 This is \nof particular interest because delays in the management of high blood pressure are \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n17 \nassociated with worse clinical outcomes.19 The clinical advisory group proposed “blood \npressure monitoring” and “QRISK2 risk scores” (or any cardiovascular risk score codes \nincluding the newer QRISK3 codes) as key measures. \nDiabetes  \nRoutine diabetes care almost entirely stopped, but largely rapidly recovered, e.g. HbA1c \ntesting was near normal by December 2020. Diabetes monitoring and foot checks remained \nslightly below normal. In some areas, concerning foot changes may be seen by specialist \nservices, hence may sometimes appear reduced in primary care. Diabetic retinopathy \nscreening recovered less well; however, specialist clinics conduct this service and send \nreports to primary care which are manually coded, therefore the sustained drop may indicate \na coding change or other provider changes. Most diabetes care activity varied widely \nbetween practices, possibly due to differences in demographics and prevalence; coding (e.g. \nuse of data entry templates in electronic health record systems); use of external providers; or \nquality of care. Our results add to the findings from earlier studies which reported a rapid \ndecline in the rate of new diabetes mellitus (DM) diagnoses and HbA1c testing in April \n2020,\n8,18 and are in line with data showing good but incomplete recovery in the following \nmonths.20,21 The clinical advisory group proposed HbA1c testing as a key sentinel measure.  \nMental Health \nMost mental health activity coded by GPs showed a sustained reduction. This was \nconsistent across various markers of activity. This was surprising given the much-discussed \nimpact of the pandemic on mental health\n22,23, but may be explained by patients either not \nseeking help or choosing other services/online resources. The latter are unlikely to explain \nall the reduction. For example, a recent study on mental health and telepsychiatry showed \nthat the rapid shift to remote service delivery has not reached some groups of patients (in \nparticular patients with dementia and mild cognitive impairment) who may require more \ntailored management24. Dementia was not widely represented in our results, perhaps being \ncovered by a range of CTv3 codes; we will conduct further research on the impacts of \nCOVID on dementia in primary care to capture this fully.  \n \nThe reduction in “Depression interim review” may warrant further investigation, but could \nreflect a change in coding behaviour. However, the similar reduction in codes for depressed \nmood would argue against this as the sole explanation. Nationally, the prescribing of \nantidepressants in primary care was sustained, indicating that access to some treatment \nwas maintained (Figure S1). Further analyses are planned before proposing any single \nmeasure for immediate ongoing monitoring, as mental health activity (especially for \ndepression and other mood disorders) spans different services such as community mental \nhealth trusts\n25, which have limited coverage in OpenSAFELY.  \n \nPrevious research similarly showed that primary care-recorded diagnosis of common mental \nhealth conditions, and associated prescribing, reduced significantly in early 2020 and did not \nrecover to pre-pandemic levels by the end of 2020.8,18,26 One region found a reduction in self \nharm in primary care sustained through to May 2021.27 Other studies suggest that the impact \non mental health may have been temporary, but following a generally worsening trend28 and \nthe English health department have responded by developing a targeted action plan.29  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n18 \nFemale and Reproductive Health \nFemale and reproductive health clinical activity generally declined modestly around April \n2020, with widespread recovery by December 2020. The reduction in contraception-related \nactivity (discussion and monitoring) was likely explained by a combination of reduced need \n(less social contact, use of non-prescription alternatives), longer repeat prescriptions, check-\nups being postponed, and long acting reversible contraception (LARC), including coils and \nimplants, not being fitted. Monthly contraceptive prescribing in England experienced only a \nsmall temporary reduction during the pandemic (\n9 and Figure S2). Six-week checks of infants \nwere well maintained, likely prioritised as vital activities, possibly aided by increased use of \ntelephone appointments and/or being carried out alongside 8-week immunisations, which \nwere also prioritised.\n30 The slight increase in breast-related symptoms by December 2020 \nmay indicate concerning delays in presentation. Current pregnancy records being slightly \nreduced may be explained by delayed presentation or increased use of self-referral directly \nto midwifery services. Other codes commonly recorded with pregnancy, e.g. date of last \nmenstrual period, would be reduced for the same reasons. The low level of gestational \ndiabetes indicates some codes for this condition were likely not captured here. Researchers \nhave previously raised concerns about disruptions to sexual and reproductive health care \nservices in the early stages of the pandemic.31 Although we observed a decline in female \nand reproductive health activities in primary care, most activities returned to near or above \npre-pandemic levels by December.  \nScreening and Related Activity \nBroadly, clinical activity related to screening declined substantially around April 2020, and \nthere was widespread recovery by December to slightly above normal levels, with the \nexception of NHS health checks. The National Screening programme paused invitations for \nbowel screening in March 2020, and they were subsequently issued at rates above 100% of \nnormal levels.\n32 NHS health checks were considered as “low priority” in the Royal College of \nGeneral Practitioners (RCGP) workload prioritisation.30 The clinical advisory group did not \npropose any measures for ongoing monitoring. \n \nSome studies outside the UK have investigated the impact caused by disruption to screening \nservices, and found evidence for example that new breast cancer diagnoses were reduced33 \nand some groups may have been affected more than others.34  \nMedication Processes  \nProcesses related to medication, in particular medication reviews, were relatively well \nmaintained during the pandemic, likely due to automated alerts commonly prompting \nclinicians when these are due. Guidance on the new Structured Medicine Reviews was \nreleased in September 2020\n35 and uptake was relatively rapid. Other related codes, such as \n“Patient understands why taking all medication” are likely recorded during structured \nmedication reviews, which explains why they also increased. Not all practices were \nrecording Structured Medicine Reviews by January 2021, likely because pharmacists with \nthe necessary training were not available in all practices. Use of this process is incentivised \nfor 2021/22 for certain patient groups.\n35 The clinical advisory group proposed a key measure \ncomprising any medication reviews.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n19 \nPolicy implications and interpretation \nThe COVID-19 pandemic brought new challenges for the NHS to deliver safe and effective \nroutine care. Despite the ongoing pressure of the pandemic, most primary care activities \nrecovered to near-normal levels by December 2020, while some aspects of routine \nmonitoring had not fully recovered. Our proposed NHS Service Restoration Observatory can \nsupport evaluation of national policies around service restoration and additionally provide \nopportunities for near real-time audit and feedback to rapidly identify and resolve concerns \naround health service activity. In particular we hope that data tools such as ours can be used \nto ensure continuity of high priority clinical services during subsequent waves of the \npandemic. \nFuture research \nAcross the clinical specialist areas we identified common themes for further research \nfollowing this study with further detail on individual areas given (detailed list in \nSupplementary Information). OpenSAFELY is a national data resource and we encourage \ninterested parties to consider exploring these patterns in the platform.  \n \n1. Monitoring activity in more granular groups  such as those with established long term \nconditions, those receiving certain medicines, or those without prior history having \nnew diagnoses, to establish the impact on each group and on new diagnoses vs \nongoing monitoring.  \n2. The level of “backlog” could be analysed to inform NHS recovery plans and to \nestablish whether those people who missed activities have later “caught up” and \nwhether there are some groups waiting longer than others.  \n3. The pandemic offers an unprecedented natural experiment in new diagnoses and \nongoing monitoring of patients’ conditions. Outcomes can be assessed to identify any \nclinical impacts on patients or tests that can be safely delayed without unintended \nimpacts to free up health care capacity. \n4. The impact on cancer referrals/stage at diagnosis in those with relevant symptoms \ne.g. breast symptoms or who missed screening. \n5. Each topic should be assessed in the context of health inequalities to explore \nwhether impacts affected some groups more than others, and should take into \naccount other activity, e.g. prescriptions, referrals, and non-primary care activity. \n \nSummary \nWe identified substantial and widespread changes in clinical activity in primary care since \nthe onset of the COVID-19 pandemic, but there was generally good recovery by December \n2020, with a few exceptions such as mental health which showed poor recovery, which \nwarrant further investigation. The authors are now further developing the OpenSAFELY NHS \nService Restoration Observatory for real-time monitoring of the key measures identified here \nas well as further studies, using primary care data to monitor and mitigate the indirect health \nimpacts of Covid-19 on the NHS. \n \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n20 \nAdministrative \n \nAcknowledgements \nWe are very grateful for all the support received from the TPP Technical Operations team \nthroughout this work, and for generous assistance from the information governance and \ndatabase teams at NHS England / NHSX. We also thank other contributors to the clinical \nadvisory group including Raj Patel, Marion Wood, Rammya Mathew, and John Geddes. \nThe views expressed are those of the authors and not necessarily those of the UK \nNational Health Service, the NIHR, or the UK Department of Health. \n \nConflicts of interest \nAuthors declare the following: BG has received research funding from the Laura and John \nArnold Foundation, the NHS National Institute for Health Research (NIHR), the NIHR \nSchool of Primary Care Research, the NIHR Oxford Biomedical Research Centre, the \nMohn-Westlake Foundation, NIHR Applied Research Collaboration Oxford and Thames \nValley, the Wellcome Trust, the Good Thinking Foundation, Health Data Research UK, the \nHealth Foundation, the World Health Organisation, UKRI, Asthma UK, the British Lung \nFoundation, and the Longitudinal Health and Wellbeing strand of the National Core Studies \nprogramme; he also receives personal income from speaking and writing for lay audiences \non the misuse of science. BMK is also employed by NHS England working on medicines \npolicy and clinical lead for primary care medicines data. AC is supported by the National \nInstitute for Health Research (NIHR) Oxford Cognitive Health Clinical Research Facility, by \nan NIHR Research Professorship (grant RP-2017-08-ST2-006), by the NIHR Oxford and \nThames Valley Applied Research Collaboration and by the NIHR Oxford Health Biomedical \nResearch Centre (grant BRC-1215-20005). AC has also received research, educational and \nconsultancy fees from INCiPiT (Italian Network for Paediatric Trials), CARIPLO Foundation \nand Angelini Pharma. SF is funded by a Wellcome Senior Research Fellowship in Clinical \nScience. RCon declares Honoraria for Work with Primary Care Pharmacy Association. KC \nworks as a GP in Oxford and is Deputy Medical Director for Primary Care, NHS England \nand NHS Improvement. MS is an NIHR funded academic clinical fellow in primary care. MM \nis NHS GIRFT Senior Clinical Advisor for Pathology. \nFunding \nThis work was jointly funded by UKRI (COV0076;MR/V015737/1), NIHR and Asthma \nUK-BLF [COV0076; MR/V015737/] and the Longitudinal Health and Wellbeing strand of \nthe National Core Studies programme (MC_PC_20030: MC_PC_20059: COV-LT-0009). \nThe OpenSAFELY data science platform is funded by the Wellcome Trust \n(222097/Z/20/Z). BG’s work on better use of data in healthcare more broadly is currently \nfunded in part by: the Bennett Foundation, the Wellcome Trust, NIHR Oxford Biomedical \nResearch Centre, NIHR Applied Research Collaboration Oxford and Thames Valley, the \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n21 \nMohn-Westlake Foundation; all Bennett Institute staff are supported by BG’s grants on this \nwork. The views expressed are those of the authors and not necessarily those of the NIHR, \nNHS England, Public Health England or the Department of Health and Social Care. Funders \nhad no role in the study design, collection, analysis, and interpretation of data; in the writing \nof the report; and in the decision to submit the article for publication.  \n \nInformation governance and ethical approval \nNHS England is the data controller; TPP is the data processor; and the key researchers on \nOpenSAFELY are acting with the approval of NHS  England. This implementation of \nOpenSAFELY is hosted within the TPP environment which is accredited to the ISO 27001 \ninformation security standard and is NHS IG Toolkit compliant;36,37 patient data has been \npseudonymised for analysis and linkage using industry standard cryptographic hashing \ntechniques; all pseudonymised datasets transmitted for linkage onto OpenSAFELY are \nencrypted; access to the platform is via a virtual private network (VPN) connection, restricted \nto a small group of researchers; the researchers hold contracts with NHS England and only \naccess the platform to initiate database queries and statistical models; all database activity is \nlogged; only aggregate statistical outputs leave the platform environment following best \npractice for anonymisation of results such as statistical disclosure control for low cell \ncounts.\n38 The OpenSAFELY research platform adheres to the obligations of the UK General \nData Protection Regulation (GDPR) and the Data Protection Act 2018. In March 2020, the \nSecretary of State for Health and Social Care used powers under the UK Health Service \n(Control of Patient Information) Regulations 2002 (COPI) to require organisations to process \nconfidential patient information for the purposes of protecting public health, providing \nhealthcare services to the public and monitoring and managing the COVID-19 outbreak and \nincidents of exposure; this sets aside the requirement for patient consent.\n39 Taken together, \nthese provide the legal bases to link patient datasets on the OpenSAFELY platform. GP \npractices, from which the primary care data are obtained, are required to share relevant \nhealth information to support the public health response to the pandemic, and have been \ninformed of the OpenSAFELY analytics platform.  \n \nThis study was approved by the Health Research Authority (REC reference 20/LO/0651).  \n \nGuarantor \nHJC is guarantor. \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n22 \nReferences \n1.  Moynihan R, Sanders S, Michaleff ZA, Scott AM, Clark J, To EJ, et al. Impact of COVID-\n19 pandemic on utilisation of healthcare services: a systematic review. BMJ Open \n[Internet]. 2021 Mar 16;11(3):e045343. Available from: \nhttp://dx.doi.org/10.1136/bmjopen-2020-045343 \n2.  NHS England. Second phase of NHS response to COVID-19 [Internet]. 2020. Available \nfrom: https://www.england.nhs.uk/coronavirus/publication/second-phase-of-nhs-\nresponse-to-covid-19-letter-from-simon-stevens-and-amanda-pritchard/ \n3.  NHS England. Urgent next steps on NHS res ponse to covid-19: Simon Stevens letter to \nlocal NHS organisations [Internet]. 2020. Available from: \nhttps://www.england.nhs.uk/coronavirus/wp-content/uploads/sites/52/2020/03/urgent-\nnext-steps-on-nhs-response-to-covid-19-letter-simon-stevens.pdf \n4.  World Health Organization. COVID /i1 19 Strategic Preparedness and Response Plan - \nOperational planning guidance to support country preparedness and response. World \nHealth Organization; 2020 May. Report No.: 1.0. \n5.  World Health Organization. Maintaining essential health services: operational guidance \nfor the COVID-19 context. World Health Organization; 2020 Jun. Report No.: \nWHO/2019-nCoV/essential_health_services/2020.2. \n6.  NHS England. Implementing phase 3 of the NHS response to the COVID-19 pandemic \n[Internet]. NHS England; 2020 Aug [cited 2021 Oct 19]. Report No.: 1.1. Available from: \nhttps://www.england.nhs.uk/publication/implementing-phase-3-of-the-nhs-response-to-\nthe-covid-19-pandemic/ \n7.  Xiong J, Lipsitz O, Nasri F, Lui LMW, Gill H, Phan L, et al. Impact of COVID-19 \npandemic on mental health in the general population: A systematic review. J Affect \nDisord [Internet]. 2020 Dec 1;277:55–64. Available from: \nhttps://www.sciencedirect.com/science/article/pii/S0165032720325891 \n8.  Williams R, Jenkins DA, Ashcroft DM, Brown B, Campbell S, Carr MJ, et al. Diagnosis \nof physical and mental health conditions in primary care during the COVID-19 \npandemic: a retrospective cohort study. Lancet Public Health [Internet]. 2020 \nOct;5(10):e543–50. Available from: http://dx.doi.org/10.1016/S2468-2667(20)30201-2 \n9.  Frazer JS, Frazer GR. Analysis of primary care prescription trends in England during the \nCOVID-19 pandemic compared against a predictive model. Fam Med Community \nHealth [Internet]. 2021 Aug;9(3). Available from: http://dx.doi.org/10.1136/fmch-2021-\n001143 \n10.  Balachandren N, Barrett G, Stephenson JM, Yasmin E, Mavrelos D, Davies M, et al. \nImpact of the SARS-CoV-2 pandemic on access to contraception and pregnancy \nintentions: a national prospective cohort study of the UK population. BMJ Sex Reprod \nHealth [Internet]. 2021 Oct 21; Available from: http://dx.doi.org/10.1136/bmjsrh-2021-\n201164 \n11.  Curtis HJ, MacKenna B, Croker R, Inglesby P, Walker AJ, Morley J, et al. OpenSAFELY \nNHS Service Restoration Observatory 1: describing trends and variation in primary care \nclinical activity for 23.3 million patients in England during the first wave of COVID-19. Br \nJ Gen Pract [Internet]. 2021 Sep; Available from: \nhttps://bjgp.org/content/early/2021/09/24/BJGP.2021.0380 \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n23 \n12.  Barron E, Bakhai C, Kar P, Weaver A, Bradley D, Ismail H, et al. Associations of type 1 \nand type 2 diabetes with COVID-19-related mortality in England: a whole-population \nstudy. The Lancet Diabetes & Endocrinology [Internet]. 2020 Oct 1;8(10):813–22. \nAvailable from: https://doi.org/10.1016/S2213-8587(20)30272-2 \n13.  Holman N, Knighton P, Kar P, O’Keefe J, Curley M, Weaver A, et al. Risk factors for \nCOVID-19-related mortality in people with type 1 and type 2 diabetes in England: a \npopulation-based cohort study. The Lancet Diabetes & Endocrinology [Internet]. 2020 \nOct 1;8(10):823–33. Available from: https://doi.org/10.1016/S2213-8587(20)30271-0 \n14.  Silverio A, Di Maio M, Citro R, Esposito L, Iuliano G, Bellino M, et al. Cardiovascular risk \nfactors and mortality in hospitalized patients with COVID-19: systematic review and \nmeta-analysis of 45 studies and 18,300 patients. BMC Cardiovasc Disord [Internet]. \n2021 Jan 7;21(1):23. Available from: http://dx.doi.org/10.1186/s12872-020-01816-3 \n15.  Williamson EJ, Walker AJ, Bhaskaran K, Bacon S, Bates C, Morton CE, et al. Factors \nassociated with COVID-19-related death using OpenSAFELY. Nature [Internet]. 2020 \nAug;584(7821):430–6. Available from: http://dx.doi.org/10.1038/s41586-020-2521-4 \n16.  NHS England and NHS Improvement. 2020/21 General Medical Services (GMS) \ncontract Quality and Outcomes Framework (QOF) - guidance for GMS contract 2020/21 \nin England [Internet]. 2020 Sep. Available from: https://www.england.nhs.uk/wp-\ncontent/uploads/2020/09/C0713-202021-General-Medical-Services-GMS-contract-\nQuality-and-Outcomes-Framework-QOF-Guidance.pdf \n17.  Fersia O, Bryant S, Nicholson R, McMeeken K, Brown C, Donaldson B, et al. The \nimpact of the COVID-19 pandemic on cardiology services. Open Heart [Internet]. 2020 \nAug;7(2). Available from: http://dx.doi.org/10.1136/openhrt-2020-001359 \n18.  Mansfield KE, Mathur R, Tazare J, Henderson AD, Mulick AR, Carreira H, et al. Indirect \nacute effects of the COVID-19 pandemic on physical and mental health in the UK: a \npopulation-based study. Lancet Digit Health [Internet]. 2021 Apr;3(4):e217–30. Available \nfrom: http://dx.doi.org/10.1016/S2589-7500(21)00017-0 \n19.  Xu W, Goldberg SI, Shubina M, Turchin A. Optimal systolic blood pressure target, time \nto intensification, and time to follow-up in treatment of hypertension: population based \nretrospective cohort study. BMJ [Internet]. 2015 Feb 5;350:h158. Available from: \nhttp://dx.doi.org/10.1136/bmj.h158 \n20.  Carr MJ, Wright AK, Leelarathna L, Thabit H, Milne N, Kanumilli N, et al. Impact of \nCOVID-19 on diagnoses, monitoring, and mortality in people with type 2 diabetes in the \nUK. Lancet Diabetes Endocrinol [Internet]. 2021 Jul;9(7):413–5. Available from: \nhttp://dx.doi.org/10.1016/S2213-8587(21)00116-9 \n21.  Carr MJ, Wright AK, Leelarathna L, Thabit H, Milne N, Kanumilli N, et al. Impact of \nCOVID-19 restrictions on diabetes health checks and prescribing for people with type 2 \ndiabetes: a UK-wide cohort study involving 618 161 people in primary care. BMJ Qual \nSaf [Internet]. 2021 Oct 12; Available from: http://dx.doi.org/10.1136/bmjqs-2021-\n013613 \n22.  COVID-19 Mental Disorders Collaborators. Global prevalence and burden of depressive \nand anxiety disorders in 204 countries and territories in 2020 due to the COVID-19 \npandemic. Lancet [Internet]. 2021 Nov 6;398(10312):1700–12. Available from: \nhttp://dx.doi.org/10.1016/S0140-6736(21)02143-7 \n23.  Pierce M, Hope H, Ford T, Hatch S, Hotopf M, John A, et al. Mental health before and \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n24 \nduring the COVID-19 pandemic: a longitudinal probability sample survey of the UK \npopulation. Lancet Psychiatry [Internet]. 2020 Oct;7(10):883–92. Available from: \nhttp://dx.doi.org/10.1016/S2215-0366(20)30308-4 \n24.  Hong JS, Sheriff R, Smith K, Tomlinson A, Saad F, Smith T, et al. Impact of COVID-19 \non telepsychiatry at the service and individual patient level across two UK NHS mental \nhealth Trusts. Evid Based Ment Health [Internet]. 2021 Nov;24(4):161–6. Available from: \nhttp://dx.doi.org/10.1136/ebmental-2021-300287 \n25.  Bakolis I, Stewart R, Baldwin D, Beenstock J, Bibby P, Broadbent M, et al. Changes in \ndaily mental health service use and mortality at the commencement and lifting of \nCOVID-19 “lockdown” policy in 10 UK sites: a regression discontinuity in time design. \nBMJ Open [Internet]. 2021 May 26;11(5):e049721. Available from: \nhttp://dx.doi.org/10.1136/bmjopen-2021-049721 \n26.  Carr MJ, Steeg S, Webb RT, Kapur N, Chew-Graham CA, Abel KM, et al. Effects of the \nCOVID-19 pandemic on primary care-recorded mental illness and self-harm episodes in \nthe UK: a population-based cohort study. The Lancet Public Health [Internet]. 2021 Feb \n1;6(2):e124–35. Available from: https://doi.org/10.1016/S2468-2667(20)30288-7 \n27.  Steeg S, Bojani\nć  L, Tilston G, Williams R, Jenkins DA, Carr MJ, et al. Temporal trends \nin primary care-recorded self-harm during and beyond the first year of the COVID-19 \npandemic: Time series analysis of electronic healthcare records for 2.8 million patients \nin the Greater Manchester Care Record. EClinicalMedicine [Internet]. 2021 \nNov;41:101175. Available from: http://dx.doi.org/10.1016/j.eclinm.2021.101175 \n28.  Gagné T, Schoon I, McMunn A, Sacker A. Mental distress among young adults in Great \nBritain: long-term trends and early changes during the COVID-19 pandemic. Soc \nPsychiatry Psychiatr Epidemiol [Internet]. 2021 Nov 11; Available from: \nhttp://dx.doi.org/10.1007/s00127-021-02194-7 \n29.  COVID-19 mental health and wellbeing recovery action plan [Internet]. GOV.UK. 2021 \n[cited 2022 May]. Available from: https://www.gov.uk/government/publications/covid-19-\nmental-health-and-wellbeing-recovery-action-plan; \n30.  RCGP. RCGP Guidance on workload prioritisation during COVID-19 [Internet]. 2021 \n[cited 2021 Nov 18]. Available from: \nhttps://web.archive.org/web/20200421140124/https://www.rcgp.org.uk/-\n/media/Files/Policy/A-Z-\npolicy/2020/covid19/RCGP%20guidance/202003233RCGPGuidanceprioritisationroutine\nworkduringCovidFINAL \n31.  Aly J, Haeger KO, Christy AY, Johnson AM. Contraception access during the COVID-19 \npandemic. Contracept Reprod Med [Internet]. 2020 Oct 8;5:17. Available from: \nhttp://dx.doi.org/10.1186/s40834-020-00114-9 \n32.  NHS England. NHS Cancer Programme: Cancer services recovery plan [Internet]. NHS \nEngland; 2020 Dec. Report No.: 001559. Available from: \nhttps://www.england.nhs.uk/coronavirus/wp-content/uploads/sites/52/2020/12/C0821-\nCOVID-19-Cancer-services-recovery-plan-14-December-2020.pdf \n33.  Eijkelboom AH, de Munck L, Lobbes MBI, van Gils CH, Wesseling J, Westenend PJ, et \nal. Impact of the suspension and restart of the Dutch breast cancer screening program \non breast cancer incidence and stage during the COVID-19 pandemic. Prev Med \n[Internet]. 2021 Oct;151:106602. Available from: \nhttp://dx.doi.org/10.1016/j.ypmed.2021.106602 \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint \n\n \n25 \n34.  Miller MM, Meneveau MO, Rochman CM, Schroen AT, Lattimore CM, Gaspard PA, et \nal. Impact of the COVID-19 pandemic on breast cancer screening volumes and patient \nscreening behaviors. Breast Cancer Res Treat [Internet]. 2021 Aug;189(1):237–46. \nAvailable from: http://dx.doi.org/10.1007/s10549-021-06252-1 \n35.  NHS England. Network Contract Directed Enhanced Service: Structured Medication \nReviews and Medicines Optimisation: Guidance [Internet]. 2020 Sep. Available from: \nhttps://www.england.nhs.uk/wp-content/uploads/2020/09/SMR-Spec-Guidance-2020-\n21-FINAL-.pdf, https://www.england.nhs.uk/publication/structured-medication-reviews-\nand-medicines-optimisation/ \n36.  BETA – Data Security Standards - NHS Digital [Internet]. NHS Digital. [cited 2020 Apr \n30]. Available from: https://digital.nhs.uk/about-nhs-digital/our-work/nhs-digital-data-and-\ntechnology-standards/framework/beta---data-security-standards \n37.  Data Security and Protection Toolkit - NHS Digital [Internet]. NHS Digital. [cited 2020 \nApr 30]. Available from: https://digital.nhs.uk/data-and-information/looking-after-\ninformation/data-security-and-information-governance/data-security-and-protection-\ntoolkit \n38.  ISB1523: Anonymisation Standard for Publishing Health and Social Care Data - NHS \nDigital [Internet]. NHS Digital. [cited 2020 Apr 30]. Available from: \nhttps://digital.nhs.uk/data-and-information/information-standards/information-standards-\nand-data-collections-including-extractions/publications-and-notifications/standards-and-\ncollections/isb1523-anonymisation-standard-for-publishing-health-and-social-care-data \n39.  Secretary of State for Health and Social Care - UK Government. Coronavirus (COVID-\n19): notification to organisations to share information [Internet]. 2020. Available from: \nhttps://web.archive.org/web/20200421171727/https://www.gov.uk/government/publicatio\nns/coronavirus-covid-19-notification-of-data-controllers-to-share-information \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted June 2, 2022. ; https://doi.org/10.1101/2022.06.01.22275674doi: medRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}