Evaluating the effect of COVID-19 on dispensing patterns: a national cohort analysis

Background Medication prescribing and dispensing often regarded as one of the most effective ways to manage and improve population health. Prescribed and dispensed medications can be monitored through data linkage for each patient. We hypothesised that changes in patient care resulting from COVID-19, changed the way patients access their prescribed medication. Objective To develop an efficient approach for evaluation of the impact of COVID-19 on drug dispensing patterns. Methods Retrospective observational study using national patient-level dispensing records in Wales-UK. Total dispensed drug items between 01-Jan-2016 and 31-Dec-2019 (counterfactual pre-COVID-19) were compared to 2020 (COVID-19 year). We compared trends of dispensed items in three main British National Formulary (BNF) sections(Cardiovascular system, Central Nervous System, Immunological & Vaccine) using European Age-Standardized rates. We developed an online tool to enable monitoring of changes in dispensing as the pandemic evolves. Result Amongst all BNF chapters, 52,357,639 items were dispensed in 2020 compared to 49,747,141 items in 2019 demonstrating a relative increase of 5.25% in 2020(95%CI[5.21,5.29]). Comparison of monthly patterns of 2020 and 2019 dispensed items showed a notable difference between the total number of dispensed drug items each month, with an average difference (D) of +290,055 and average Relative Change (RC) of +5.52%. The greatest RC was observed in a substantial March-2020 increase (D=+1,501,242 and RC=+28%), followed by second peak in June (D=+565,004, RC=+10.97%). May was characterised by lower dispensing (D=-399,244, RC=-5.9%). Cardiovascular categories were characterised, across all age groups, by dramatic March-2020 increases, at the epidemic peak, followed by months of lower than expected dispensing, and gradual recovery by September. The Central Nervous System category was similar, but with only a short decline in May, and quicker recovery. A stand-out grouping was Immunological and Vaccine, which dropped to very low levels across all age groups, and all months (including the March dispensing peak). Conclusions Aberration in clinical service delivery during COVID-19 led to substantial changes in community pharmacy drug dispensing. This change may contribute to a long-term burden of COVID-19, raising the importance of a comprehensive and timely monitoring of changes for evaluation of the potential impact on clinical care and outcomes

399,244, RC=-5.9%). Cardiovascular categories were characterised, across all age groups, by 48 dramatic March-2020 increases, at the epidemic peak, followed by months of lower than 49 expected dispensing, and gradual recovery by September. The Central Nervous System 50 category was similar, but with only a short decline in May, and quicker recovery. A stand-out 51 grouping was Immunological and Vaccine, which dropped to very low levels across all age 52 groups, and all months (including the March dispensing peak). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 19, 2021.

65
The novel coronavirus disease 2019 (COVID-19) pandemic has resulted in unprecedented 66 changes on health care services provision (1). While technology has provided an increase in 67 telephone or virtual appointments there has been an overall net reduction in primary care 68 appointments (2). Increase in demand for essential medicines coupled with complex medicines 69 supply chain issues, enforcement of social distancing, quarantine, and self-isolation has 70 impacted the prescribing and dispensing of medicines (3)(4). 71 Studies using survey and patient data, reported an observed change in medication use during 72 the pandemic (5)(6). Changes in service delivery have been monitored through multiple 73 national audits which often lag considerably behind in real-time (7)(8). There is an urgent need 74 to focus appropriate resources on optimisation of medication dispensing monitoring systems, 75 especially in vulnerable patient groups; for example those receiving immunosuppressive drugs 76 (9).

77
Electronic dispensing records holding information on all dispensed prescriptions in primary 78 care, provide a unique opportunity to monitor dispensing trends. We aimed: 1) to measure the 79 general impact of COVID-19 on dispensing patterns; 2) create a national research ready dataset 80 of all primary care dispensing records for the whole population of Wales and 3) to design and 81 implement a tool for monitoring and visualisation real-time trends.

83
Study design and data sources 84 We conducted a retrospective observational study, accessing national patient-level primary 85 care dispensing data using the SAIL Databank (10)(11). SAIL holds the Welsh Dispensing is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint   Products & Vaccines. We calculated age-standardised dispensing rates per 100,000 population.   Interactive tool 131 We developed an interactive tool using R shiny (18) which allows dynamic monitoring of 132 trends in the current year and enables comparison between the most up to date data with 133 previous years' patterns. The dashboard is accessible from https://wdds.ml/. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 19, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint        Our data showed that this peak occurred in most drug categories and was followed by a dip in 216 the following two months (lockdown period) which is suggestive of batch-dispensing in 217 preparation of the national lockdowns. It is also likely that patients may have had spare 218 prescriptions to hand and felt an urgent to get them dispensed. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The patient level data used in our study are available as a monthly extract to SAIL Databank 221 at the patient level; this provides the opportunity to use aggregated level data for monitoring 222 trends in near real-time, as well as exploring the effect of patient-level factors such as age, sex, 223 socioeconomic status and ethnicity. This mechanism provides the powerful opportunity to 224 using these data not only to monitor and evaluate prescribing and dispensing trends in near 225 real-time, but also to explore the effect of changing treatment patterns on outcomes considering  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 19, 2021. ; https://doi.org/10.1101/2021.02.15.21251552 doi: medRxiv preprint 18 | P a g e appropriate use of SAIL data. When access has been approved, it is gained through a privacy-245 protecting safe haven and remote access system referred to as the SAIL Gateway. SAIL has 246 established an application process to be followed by anyone who would like to access data via 247 SAIL https://www.saildatabank.com/application-process. This study has been approved by  The main patient-level data sources used in this study are available in the SAIL Databank at 257 Swansea University, Swansea, UK, but as restrictions apply they are not publicly available. All 258 proposals to use SAIL data are subject to review by an independent Information Governance 259 Review Panel (IGRP). Before any data can be accessed, approval must be given by the IGRP.

260
The IGRP gives careful consideration to each project to ensure proper and appropriate use of 261 SAIL data. When access has been granted, it is gained through a privacy protecting safe haven 262 and remote access system referred to as the SAIL Gateway. SAIL has established an application 263 process to be followed by anyone who would like to access data via SAIL at 264 https://www.saildatabank.com/application-process/ 265

Competing interests 266
The author(s) declare(s) that they have no competing interests. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 19, 2021. ; https://doi.org/10.1101/2021.02.15.21251552 doi: medRxiv preprint 20 | P a g e This work uses data provided by patients and collected by the NHS as part of their care and 289 support. We would also like to acknowledge all data providers who make anonymised data 290 available for research. 291 We wish to acknowledge the collaborative partnership that enabled acquisition and access to 292 the de-identified data, which led to this output. The collaboration was led by the Swansea