EBQ:Relevance of Discharge Glucose Levels

Revision as of 12:19, 1 February 2017 by Rossdonaldson1 (talk | contribs) (See Also)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Under Review Journal Club Article
Driver B, Olives T, Bischof J, Salmon M, Miner J. "Discharge Glucose is Not Associated With Short-Term Adverse Outcomes in Emergency Department Patients With Moderate to Severe Hyperglycemia". Annals of Emergency Medicine. 2016. 68(6):697-705.
PubMed Full text PDF

Clinical Question

Is the ED discharge glucose level associated with adverse outcomes?

Conclusion

Higher discharge glucose levels were not associated with a greater risk of repeated ED visits, hospitalization, or other adverse outcomes.

Major Points

Elevated blood glucose levels have long been known to be associated poor long term outcomes, as well as short term complications of critically ill and hospitalized patients. ED physicians, however, have long struggled with the appropriate management of moderate-severe hyperglycemia in patients appropriate to discharge from the ED. Up to this point, there are no guidelines for a "safe" discharge glucose level, or if the finding is associated with significant short term adverse outcomes. In this retrospective chart review, the authors looked at Type II diabetics with glucose readings > 400 and discharged from the ED, and sought to verify if there was an association between higher glucose levels and poor 7 day outcomes. The authors concluded that no association exists, and that attaining a specific glucose level prior to discharge in patients not requiring admission is likely less important that previously thought.

Study Design

This was a retrospective cohort chart review that searched for patients with glucose levels > 400 and discharged from the ED. Outcome measures were obtained by again searching the electronic records within the study network, as well as ambulance registry. The study took place at an urban Level I trauma center with an average ED census of 100,000 patients. The study took place from January 1, 2010 through December 31, 2011.

Population

Patient Demographics

  • 58% men, 42% women
  • Mean age = 46
  • Most common chief complaint was hyperglycemia (49%), followed by alcohol intoxication (11%)
  • Mean arrival glucose = 491

Inclusion Criteria

All patients who were over the age of 18 and had a glucose reading at any type of 400 or greater were evaluated for inclusion.

Exclusion Criteria

The following patients were excluded from the study:

  • patients admitted to the hospital
  • those known to be Type I diabetics
  • if the chief complaint was hypoglycemia

Interventions

No specific interventions were mandated in the study, and the outcomes were compared to discharge glucose absolute values, not interventions. However, the authors did include information on interventions taken to lower glucose levels while in the ED. Specifically:

  • 60% of patients received both IV fluids and SubQ insulin
  • 12% of patients received neither IV fluids or insulin.

Outcomes

  • With 13% of patients lost to follow up, 495 encounters remained for analysis.
  • Mean discharge glucose in analyzed patients ranged from 48-694.

Primary Outcome

The Primary Outcome looked to determine if there was an association between discharge glucose and 7 day adverse outcomes (return ED visit for hyperglycemia, or hospitalization).

  • Return ED visits for hyperglycemia: 13% of patients (n=62)
  • Patients requiring hospitalization: 7% (n=36)
  • DKA was rare, occurring in only 2 patients
  • No deaths were observed
  • Mean discharge glucose for patients WITH and WITHOUT 7 day adverse events was 317 mg/dL and 336 mg/dL respectively. After adjusting for confounders, no associated between discharge glucose and repeat ED visit or hospitalization was seen.

Secondary Outcomes

A secondary dichotomous variable model was constructed to evaluate if a discharge glucose level greater than 350 mg/dL was associated with greater 7 day adverse effects; no such affect was determined.

Criticisms & Further Discussion

  • Type I diabetes were excluded from this study, so the results should not be applied to these patients
  • Retrospective
  • Potential outcome bias, based upon the fact that patients may seek care in outside network hospitals and systems, which would not allow for the true 7 day outcomes to be measured

External Links

See Also