Abstract

Aim: To compare the impact and effectiveness of introducing reflective and reflex testing of magnesium in severe hypokalaemia.

Methods: All specimens with [K] ⩽2.5 mmol/l were retrospectively identified in three 6-month periods: baseline, with reflective testing, and with reflex testing. For each episode of hypokalaemia it was noted whether [Mg] was measured.

Results: Measurement of [Mg] increased from 7.7% to 63.9% (p<0.001) after introducing reflective testing, and then to 98.7% (p<0.001) with reflex testing. Diagnosis of hypomagnesaemia increased from 7.7% to 43.1% (p<0.001) and 69.3% (p<0.01) with reflective and reflex testing, respectively. For severe hypomagnesaemia ([Mg] <0.50 mmol/l) the increase was from 1.9% to 8.3% (p = 0.127 relative to baseline) with reflective testing and then to 12.0% with reflex testing (p<0.05 relative to baseline and p = 0.463 relative to reflective testing). The number of tests needed to diagnose was similar for reflective and reflex testing: 1.48 and 1.42 for hypomagnesaemia, respectively; and 7.67 and 8.22 for severe hypomagnesaemia, respectively. 42 and 70 extra magnesium assays compared to baseline were requested due to reflective and reflex testing, respectively.

Conclusion: Reflex testing was the most time-efficient and consistent method of diagnosing hypomagnesaemia in severe hypokalaemia. This was mainly due to the increased number of magnesium assays performed. However, as the absolute increase in test numbers was small (28 in a 6-month period) and the test is inexpensive, selective reflex testing can improve quality in a cost-efficient manner.