Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Arterial or venous blood gas estimation for monitoring and risk stratification following tricyclic antidepressant overdose

Three Part Question

In [patients who have taken an overdose of tricyclic antidepressants] does [measurement of arterial or venous blood gases] lead to [superior risk stratification and monitoring of blood pH]?

Clinical Scenario

A tearful eighteen year-old lady presents to the Emergency Department claiming to have "stupidly" taken thirty of her dothiepin tablets after a bottle of wine. She is extremely needle phobic and you site an intravenous cannula only after exercising your well honed skills of persuasion and distraction. You know that blood gas analysis is an integral part the initial assessment but wonder whether a venous sample would be as effective as an arterial sample for risk stratification and monitoring in this situation.

Search Strategy

Ovid Medline 1950 - 2008 June Week 1
Ovid Embase 1980 - 2008 Week 23
(exp Antidepressive Agents, Tricyclic/ OR OR OR exp Amitriptyline/ OR OR exp Desipramine/ OR OR exp Clomipramine/ OR OR exp Doxepin/ OR OR exp Dothiepin/ OR OR exp Imipramine/ OR OR exp Lofepramine/ OR OR exp Nortriptyline/ OR OR exp Trimipramine/) AND (Exp Blood Gas Analysis/ OR exp Blood Gas/ OR blood gas$.mp.) limit to human and English language

Search Outcome

A total of 81 papers were identified (65 in Medline, 18 in Embase). One paper was directly relevant to the three-part question.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Eizadi-Mood et al
50 patients with clinical manifestations of tricyclic antidepressant poisoning who presented to the Emergency Department. Samples for arterial and venous gas analysis were obtained at presentation and 30 minutes after bolus sodium bicarbonate therapy.Prospective diagnostic cohort studyMean (SD) pH on admissionVenous 7.34 (0.0049); Arterial 7.37 (0.0052). P=0.00Small statistically significant difference sin parameters identified but clinical significance of the difference in parameters not assessed. No attempt to correlate blood gas parameters with the incidence of complications.
Mean (SD) HCO3 on admissionVenous 23.26 (3.23); Arterial 22.19 (3.28). P=0.01
Mean (SD) PO2 on admissionVenous 42.50 (10.78); Arterial 79.94 (15.94). P=0.00
Mean (SD) PCO2 on admissionVenous 43.79 (6.39); Arterial 38.47 (7.10). P=0.00
Mean (SD) pH 30 min after bicarbonateVenous 7.34 (0.049); Arterial 7.37 (0.042). P=0.12
MEan (SD) HCO3 30 min after bicarbonateVenous 25.24 (3.35); Arterial 23.78 (3.11). P=0.23
Linear regression model (arterial and venous pH measurements)Significant relationship (P<0.001). r squared = 0.60


Assessment of acid-base balance is an essential part of the initial assessment and monitoring of patients who have taken a significant overdose of tricyclic antidepressants. An important part of the management of these patients is alkalinisation, which has been reported to result in profound alkalaemia and high mortality (Wrenn et al, 1992). However arterial blood sampling is often painful. In alert patients who do not have suspected hypoventilation, venous blood gas analysis would be preferable if it could be shown to be equivalent for risk stratification and monitoring. The only relevant paper did seek to directly answer this question and had an appropriate sample size. Although statistically significant differences were detected in all relevant parameters between arterial and venous blood gas analysis, the clinical effects of the differences in bicarbonate and pH (in particular) are questionable. Further, a fairly strong linear relationship was demonstrated between arterial and venous pH measurements. The study did not attempt to determine which sampling method enabled superior prediction of complications. However the evidence is sufficient to recommend that venous blood gas analysis is likely to be acceptable for the initial assessment and subsequent monitoring of these patients, so long as hypoxia or hypoventilation are not suspected.

Clinical Bottom Line

Venous blood gas analysis is an acceptable alternative to arterial blood gas analysis following tricyclic antidepressant overdose unless hypoxia or hypoventilation are suspected.

Level of Evidence

Level 3 - Small numbers of small studies or great heterogeneity or very different population.


  1. Wrenn K; Smith BA; Slovis CM Profound alkalemia during treatment of tricyclic drug overdose: A potential hazard of combined hyperventilation and intravenous bicarbonate American Journal of Emergency Medicine 1992; 10(6): 553-555
  2. Eizadi-Mood N; Moein N; Saghaei M Evaluation of relationship between arterial and venous blood gas values in the patients with tricyclic antidepressant poisoning Clinical Toxicology 2005; 43: 357-360