<sub>Urgent need for alternative sampling strategies for electrolytes in primary care: relationship between potassium in primary care samples and ambient temperature on day of collection in a population of 600,000 registered patients.</sub>
The problem: Potassium is a commonly sampled electrolyte in general practice and results guide management for patients on diuretic therapy, ace inhibitors/angiotensin receptor blockers, digoxin and spironolactone. This affects patients with common long term conditions such as cardiovascular disease, diabetes, hypertension and heart failure. Reports from rural and remote general practices have described a relationship between potassium results for blood samples collected in general practice and the ambient temperature on the day of collection. However we do not know if this is the case in urban and semi-rural areas where the majority of the population live. We tested the hypothesis that ambient temperature is a determinant of primary care potassium results in urban and semi rural areas in Oxfordshire general practices.The approach: All electrolyte results from samples taken in 87 Oxfordshire general practices in patients > 18 years between 01/2009 and 12/2010 were included (total registered population 600,000). Blood was sent in lithium/heparin tubes to the chemistry laboratories of the Oxford University Hospitals NHS Trust using routine transportation and storage. Daily temperatures in degrees Celsius for Oxfordshire were taken from the Meteorological Office national temperature surveillance system. The relationship between hypokalaemia (potassium < 3.5mmol/L) and temperature was tested with logistic regression adjusting for age and distance from laboratory using SPSS version 20.
184,858 blood samples were taken in 111,248 patients. Among all results, 10% showed hypokalaemia (K < 3.5.mmol/L) and 4.7% showed hyperkalaemia (K>5.0 mmol/L). There was a strong negative correlation between potassium level and daily temperature when the blood was taken (Pearson R -0.246, P<0.001) indicating lower potassium at higher temperatures. Prevalence of hypokalaemia ranged from 1% (daily temperature -8oC) to 25% (daily temperature 20oC). Multivariable logistic regression of hypokalaemia prevalence against daily temperature, adjusting for patient age and distance from practice to hospital laboratory, showed a highly significant negative relationship (Wald test 47695.6, P< 0.001).Consquences: Potassium results from primary care samples vary with ambient temperature in an urban and semi- rural population, with significant impact on the diagnosis of hypokalaemia. The probability that a patient will be diagnosed with hypokalaemia therefore varies significantly with temperature on day of collection. Patients with a normal potassium are likely to be inaccurately diagnosed with hypokalaemia on warmer days which can alter decisions on drug treatment and even affect referrals to hospital. Importantly, patients with high potassium, which is most often caused or exacerbated by medication may be falsely reassured by their GP that they have a normal potassium in warmer weather, leaving medication unaltered. Alternative strategies to improve the accuracy of potassium sampling in primary care are needed in order to provide routine healthcare, including provision of near patient testing to avoid errors due to transportation.
- Thanusha Ananthakumar, Department of Clinical Biochemistry, Oxford University Hospitals NHS Trust, Oxford, Oxfordshire, UK
- Richard Clark, Department of Clinical Biochemistry, Oxford University Hospitals NHS Trust, Oxford, Oxfordshire, UK
- Brian Shine, Nuffield Department of Medicine, University of Oxford, Oxford, Oxfordshire, UK
- Chris O'Callaghan, NIHR Oxford Biomedical Research Centre, Oxford, Oxfordshire, UK
- Dan Lasserson, Department of Clinical Biochemistry, Oxford University Hospitals NHS Trust, Oxford, Oxfordshire, UK