Medical montage concept

Medicines safety

How can we identify patients who are most at risk of preventable medication-related problems?

The work described here was presented at the Pharmacy Together 2019 conference.

There is growing evidence of the need to improve medication safety, as outlined by the World Health Organization’s Global Patient Safety Challenge: Medication Without Harm. Clinical pharmacy plays a key role in improving medication safety but increasing demands on hospital pharmacy services have resulted in calls for clinical prioritisation. This study aims to inform prioritisation by quantifying associations between clinically relevantmedication-related problems (MRPs) and potential risk factors and groups. 

The objective of this work was to quantify statistical associations between risk factors and the study outcome: preventable MRPs that were at least moderate in severity.

Patients from adult medical wards at two UK hospitals were prospectively included into this observational study between April and November 2016. Eighteen potential risk factors were pre-selected based on previous research and expert opinion. Univariable and multivariable logistic regression modelling was used to determine relationships between risk factors and the study outcome. This study required and received ethics approval.

Among 1,503 eligible admissions, 12 of the 18 risk factors/groups were associated with the study outcome on univariable analysis: 

  • age
  • number of comorbidities
  • socioeconomic status
  • previous allergy
  • number of previous hospital admissions
  • number of medicines prescribed
  • parenteral medicine administration
  • renal function
  • serum albumin
  • white cell count
  • use of high-risk medicines (specifically systemic antimicrobials, antidepressants, anticoagulants, anti-diabetic medication, epilepsy medicines and aminoglycosides/glycopeptides)
  • primary diagnosis (genitourinary and musculoskeletal-integumentary systems).

Body mass index, history of dementia, liver disease, serum potassium, serum sodium and platelet count were not associated with the study outcome. 

Multivariable analysis found only five risk factor/groups to be independently associated with the outcome: 

  • number of comorbidities (odds ratio [OR] 1.15, 95% confidence interval [CI] 1.07-1.22)
  • socioeconomic status (OR 1.05, 95% CI 1.01-1.09)
  • previous allergy (OR 1.30, 95% CI 1.03-1.64)
  • use of high-risk medicines (epilepsy medicines and systemic antimicrobials, OR 1.61, 95% CI 1.16-2.25 and OR 1.44, 95% CI 1.08-1.92 respectively)
  • and the primary diagnosis ‘gastrointestinal system’ (OR 0.52, 95% CI 0.32-0.86). 

The majority of adjusted ORs were also lower compared to the equivalent univariable analyses.

Multivariable analysis suggests that associations between the study outcome and risk factors/groups can be explained, partly or fully, by other risk factors. This suggests that risk is multifactorial; a combination of risk factors may therefore need to be considered when developing clinical prioritisation tools.

Strengths of this research include adherence to prognostic research recommendations; this has potential to enhance quality of data collection and reduce bias. Other strengths include the large sample size and use of two study sites to increase generalisability. A limitation of the study is possible underestimation of the prevalence of MRPs due to the observational nature of the study.


This work was presented as a poster at the Pharmacy Together 2019 conference, organised by UKCPA and Pharmacy Management.

Additional authors:

Li Wei and Bryony Dean Franklin, UCL School of Pharmacy, London.

The opinions expressed in this article are those of the author. They do not purport to reflect the opinions or views of the UKCPA or its members. We encourage readers to follow links and references to primary research papers and guidance.

Competing interest statement: 

The author declares: this work was funded by the National Institute for Health Research; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

Subscribe

Get the latest clinical pharmacy articles direct to your inbox