Project objectives

 The original objectives of this work were to:

1. Review the literature to ascertain clinical outcomes, health status and resource use data to populate new economic models for different NIHR/RCGP prescribing safety indicators;
2. Develop up to six new economic models based on different NIHR SPCR/RCGP prescribing safety indicators to identify those indicators that are likely to be the most cost-effective.

Changes to project objectives

 There have been no changes to the project originally outlined in the proposal. However, we had to request a no-cost extension to allow us to finish all six economic models. The extra time was required to build those models for which there are no existing similar models in the literature. Building these models requires extensive literature searching, clinical input and team discussions. Due to the unanticipated volume of work and key staff changes, this work is still ongoing. 

Brief summary 

The focus of this project is on developing up to six economic models needed for the proposed economic evaluation of the roll-out of the PINCER intervention. The models focus on the use of non-steroidal anti-inflammatory drugs (NSAIDs) and antiplatelet drugs:

1. Prescription of an NSAID in a patient with heart failure;
2. Concurrent use of two NSAIDS for more than two weeks (not including low dose aspirin);
3. Aspirin or clopidogrel prescribed to people with previous peptic ulcer or GI bleed without gastro-protection (2 models);
4. Prescription of an NSAID in a patient with chronic renal failure with an eGFR <45;
5. Prescription of warfarin and aspirin in combination (without co-prescription of gastroprotection);
6. Prescription of warfarin in combination with an oral NSAID.

This has involved:

1. Conducting a literature search of the electronic databases Medline, Embase and Web of Science using the treatment pathway specific search terms. Full texts of the retrieved references have then been evaluated. Finally, reference lists of the retrieved references have been hand-searched;
2. Taking transition probability and health status data, preferentially from up-to-date UK sources that reflect the characteristics of the populations seen within UK primary care;
3. Undertaking the preparatory work needed to describe secular trends in rates of hazardous prescribing and potentially inadequate blood-test monitoring for these indicators across a large number of practices contributing to the Clinical Practice Research Datalink, along with intra-cluster correlation coefficients to help undertake sample size calculations. We have also undertaken a series of cohort studies in collaboration with colleagues in Epidemiology and Public Health to determine the risks of upper gastrointestinal bleeding while being prescribed NSAIDs or aspirin;
4. Developing six new economic models based on different NIHR SPCR/RCGP prescribing safety indicators. This has involved developing a Markov model for each treatment pathway, using a five year time horizon. Each treatment pathway describes the consequences of being prescribed or monitored appropriately, compared with being prescribed or monitored inappropriately;
5. Discussing each model with clinicians on the PINCER team and clinical experts in the area to ensure face validity;
6. Populating each error-specific model with probability, cost and health status data. This has allowed the generation of the outcomes and costs in a cohort of patients with the error present, and in a cohort with the error absent. The probability, cost and utility data will be assigned beta, gamma and beta distributions respectively;
7. Calculating the incremental cost per extra QALY generated in the absence of an error, using the following equation: 
   (Cost^{error absent}– Cost^{error present}) / (QALY^{error absent} – QALY^{error present})

For a future large-scale roll-out of the PINCER intervention, we will then be able to combine the treatment pathway models with the data collected from the trial. This will allow us to generate probabilistic cost per QALY and net benefit statistics to assess how cost-effective our intervention is when translated into main-stream general practice.

We have now completed the first two economic models. This work is ongoing; we are in the final stages of completing both the aspirin and clopidogrel models and have begun reviewing the literature for the remaining models.

This funding has not covered the costs of the intervention (which for the PINCER rollout across the East Midlands are being met by the Health Foundation and/or Clinical Commissioning Groups with additional funding from the East Midlands Academic Health Science Network). However, the funding that we have received from the NIHR School for Primary Care Research has allowed us to do the necessary work to apply for an NIHR programme grant. We submitted a second stage application in October 2015 and will know if this has been successful in March 2016. 

Plain English summary

We have completed a large study that has shown that the number of medication errors in general practices can be reduced if a) patients at risk are identified from GP computer systems and b) pharmacists work with GPs to make the prescribing or medication monitoring safer. The study was called the PINCER trial and the results were published in The Lancet in 2012.

The Department of Health in England, and a number of Clinical Commissioning Groups in general practice, are interested in rolling out the methods we used in the PINCER trial to large numbers of general practices. Not only might this approach help prevent unnecessary harm to patients, but it may also reduce the costs associated with dealing with prescribing errors, which sometimes require hospital admission.

We want to be able to be in a position to find out how cost-effective any roll-out is likely to be, compared with the findings from the original study, as it is important to know if our intervention has the potential to make cost savings to the NHS in addition to improving patient safety. In order to do this, we need to explore the benefits and potential cost savings involved in identifying and addressing different types of prescribing errors so that those most likely to benefit patients and reduce costs are included in the rollout of the methods used in the PINCER trial.

In order to answer these questions, and estimate the likely costs to the NHS, we need to apply for further funding to carry out a much bigger research study. The funding that we have received from the NIHR School for Primary Care Research has allowed us to do the necessary work to apply for an NIHR programme grant which we submitted in October 2015.

This has included reviewing the research literature to develop economic models for the prescribing of a limited range of medications that are most commonly associated with medication error. 

Dissemination

Published articles

1. Stocks SJ, Kontopantelis E, Akbarov A, Rodgers S, Avery AJ, Ashcroft DM. Examining variations in prescribing safety in UK general practice: a cross-sectional study using the Clinical Practice Research Datalink. British Medical Journal 2015;351:h5501.
   http://www.bmj.com/content/351/bmj.h5501
2. Akbarov A, Kontopantelis E, Sperrin M, Stocks SJ, Williams R, Rodgers S, Avery A, Buchan I,  Ashcroft DM. Primary care medication safety surveillance with integrated primary and secondary care electronic health records: a cross-sectional study. Pharmacoepidemiology & Drug Safety 2015;38(7):671-82 DOI:10.1007/s40264-015-0304-x.
   http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4486763/
3. Elliott RA, Putman KD, Franklin M, Annemans L, Verhaeghe N, Eden M, Hayre J, Rodgers S, Sheikh A, Avery AJ (on behalf of the PINCER Team). Cost-effectiveness of a pharmacist-led information technology intervention for reducing rates of clinically important errors in medicines management in general practices (PINCER). PharmacoEconomics 2014;32(6):573-590 doi: 10.1007/s40273-014-0148-8.
   http://link.springer.com/article/10.1007%2Fs40273-014-0148-8
4. Avery AJ, Rodgers S, Franklin BD, Elliott RA, Howard R, Slight SP, Swanwick G, Knox R, Gookey G, Barber N, Sheikh A. Research into practice: safe prescribing. British Journal of General Practice 2014;64(622):259-61 doi: 10.3399/bjgp14X679895.
   http://bjgp.org/content/64/622/259
5. Spencer R, Bell B, Avery AJ, Gookey G, Campbell SM. Identification of an updated set of prescribing safety indicators for GPs. British Journal of General Practice 2014;64 (621);e181-e190.
   http://bjgp.org/content/64/621/e181
6. Sadler S, Rodgers S, Howard R, Morris CJ, Avery AJ (on behalf of the PINCER triallists). Training pharmacists to deliver a complex information technology intervention (PINCER) using the principles of educational outreach and root cause analysis. International Journal of Pharmacy Practice 2014;22:47-58.
   http://onlinelibrary.wiley.com/doi/10.1111/ijpp.12032/full
7. Howard R, Rodgers S, Avery AJ, Sheikh S (on behalf of the PINCER triallists). Description and process evaluation of pharmacists’ interventions in a pharmacist-led information technology-enabled multicentre cluster randomised controlled trial for reducing medication errors in general practice (PINCER trial). International Journal of Pharmacy Practice 2014;22:59-68.
   http://onlinelibrary.wiley.com/doi/10.1111/ijpp.12039/full

Reports to funding bodies

1. Elliott RA, Putman K, Franklin M, VerhaegheN, Annemans L, Eden M, Hayre J, Rodgers S, Cantrill JA, Armstrong S, Cresswell K, Hippisley-Cox J, Howard R, Kendrick D, Morris CJ, Murray SA, Prescott RJ, Swanwick G, Boyd M, Tuersley L, Turner T, Vinogradova Y, Sheikh A, Avery AJ. Economic evaluation of a pharmacist-led IT-based intervention with simple feedback in reducing rates of clinically important errors in medicines management in general practices (PINCER). Report to the Department of Health National Patient Safety Research Portfolio, February 2013.
   http://www.birmingham.ac.uk/Documents/college-mds/haps/projects/cfhep/psrp/finalreports/PINCER-economics-report.pdf

Peer reviewed conference presentations

1. Stocks SJ, Kontopantelis E, Akbarov A, Rodgers S, Anthony AJ, Ashcroft D. Examining variations in prescribing safety in UK general practice: a cross-sectional study using the Clinical Practice Research Datalink. Oral presentation, Society for Social Medicine 59th Annual Scientific Meeting, 2 September 2015, Dublin.
2. Rodgers S, Avery AJ. Development of prescribing safety indicators and data extraction methods in UK general practice. Poster presentation, 42nd North American Primary Care Research Group Annual Meeting, 21-25 November 2014, New York.
3. Rodgers S, Avery AJ. Development of prescribing safety indicators and data extraction methods in UK general practice. Poster presentation, NIHR School for Primary Care Research Showcase Event, 26 September 2014, Oxford.
4. Stocks, SJ, Kontopantelis E, Rodgers S, Avery AJ, Ashcroft D. Application of Prescribing Safety Indicators to the Clinical Practice Research Database: a retrospective cohort study. Oral presentation, 43rd Annual Scientific Meeting of the Society of Academic Primary Care, 9-11 July 2014, Edinburgh.
5. Rodgers S, Avery AJ, Evans D, Morrell T and the PRIMIS Training and Informatics Teams. Development of prescribing safety indicators and data extraction methods in UK general practice. Poster presentation, 42nd Annual Scientific Meeting of the Society of Academic Primary Care, 3-5 July 2013, Nottingham.

The project is ongoing and there are plans to publish the economic models in peer-reviewed journals.

Public involvement

Mr Glen Swanwick is a co-applicant on this bid and has been providing a public and patient perspective on the project. Mr Swanwick was one of the public and patient representatives on our PINCER trial and indeed was a co-author on the Lancet publication that presented the findings from the trial. In the current project, he has been providing particular advice on how members of the public might perceive the economic evaluation of the roll-out of the PINCER trial intervention in general practices across the country.

Impact

Once we have developed the economic models we will be in a position to include those prescribing indicators most likely to be cost effective in the next wave of the PRIMIS PINCER Query Library for national rollout. We will also be in a position to incorporate the findings from the economic models into CHART online software to enable GP practices and CCGs to view potential cost savings that could be made if numbers of patients at risk of medication error are reduced. As a result of this work, we will be able to undertake an economic evaluation of any large-scale roll-out of the PINCER trial methods.

This project was funded by the National Institute for Health Research School for Primary Care Research (project number 203)

Department of Health Disclaimer

The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the NIHR School for Primary Care Research, NIHR, NHS or the Department of Health.