Hypoglycaemia: could be reliably diagnosed with reagent strips

Clinical bottom line (level 2b)

  1. In patients who required glucose determination in the emergency room, reagent strips could reliably diagnose hypoglycaemia (LR-0.04) , (LR+16)
Scott et al: Ann Emerg Med 1998; 32: 305-309
Expires December 2003

The study

Setting: University hospital emergency department in Northern USA

211 patients (aged ?, ?% male) 'requirement' for electrolyte panel including glucose

Excluded if
  • post-hoc exclusion of 4 patients with glucose beyond the range for the reagent strips (>800 mg/dL)



    • Independent blinded reference standard, applied in ?all patients from a non-consecutive ?appropriate spectrum.
    Reference standard:
    • laboratory glucose value of <60 mg/dL measured on a simultaneous blood sample
    Diagnostic test: Chemstrip bG reagent strip reagent strip with a reading of 60mg/dL or below considered hypoglycaemic
    • visual reading from reagent strip alone: no reflectance meters used
    • no additional or intensified training was given in the use of the reagent strips

    The evidence

    pre-test probability of hypoglycaemia: 14%, (95% CI: 11% to 18%)

    diagnostic test laboratory glucose <60 mg/dL reagent strip <60 mg/dL LR+
    (95% CI)    		 post-test probability LR-
    (95% CI)    		 post-test probability
    28 11 16
    (9.0 to 28)     		72% 0.04
    (0.01 to 0.25)     		1%
    total 29 182

    Comments

    1. The study demonstrates that test glucose values greater than 60 mg/dl by reagent strips are highly likely to be a true positive result and, for pre-test probabilities less than 50%, effectively rules out hypoglycaemia. Moreover, hypoglycaemia by reagent strip would, for pre-test probabilities of 50% or greater, rule in the diagnosis.
    2. 'Convenience sample' not further defined - uncertain how appropriate the sample is
    3. 4 exclusions for hyperglycaemia beyond range of reagent strip probably do not change the validity of the results
    4. No information on total number of analyses carried out in laboratory or patients seen over the study period; are we seeing a representative sample?
    5. Results similar to those described previously for use in the prehospital setting.

    Citation

    1. Scott PA, Wolf LR, Spadafora MP: Accuracy of reagent strips in detecting hypoglycemia in the emergency department.. Ann Emerg Med 1998; 32: 305-309
    Contributor: MJ Taylor and Bob Phillips, January 1998
    Reviewer: Guy De Bruyn

    Clinical Question.
    Patient in patients presenting to the emergency room
    Intervention or Exposure are reagent strips
    Outcome able to reliably diagnose hypoglycaemia