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

Doxapram resurrected? Weak evidence of efficacy suggests a potential use in patients ineligible for non-invasive ventilation

Three Part Question

In [patients with an acute exacerbation of COPD who are unwilling or unable to have ventilatory support] is [administration of doxapram or standard therapy alone] associated with [an improved outcome and recovery] in the acute setting?

Clinical Scenario

A 73-year-old male with known chronic obstructive pulmonary disease (COPD) presents to the emergency department with dyspnoea. He is diagnosed with an acute exacerbation of COPD. Despite optimal medical management he fails to improve, his arterial blood gases demonstrate type II respiratory failure with worsening respiratory acidosis. He is offered a trial of non-invasive ventilation but refuses as. A decision is taken that invasive ventilation is not in the best interests of the patient. You wonder if the respiratory stimulant doxapram may improve the patient's respiratory failure.

Search Strategy

We searched MEDLINE 1948 - October Week 4, 2011; EMBASE database 1980 - 2011 Week 44, using the OVID interface, and the Cochrane Database of Systematic Reviews. We also reviewed the reference lists of relevant papers for additional papers that addressed the three-part question.

(exp doxapram/OR OR AND (exp Chronic Obstructive Lung Disease/ OR exp Chronic Bronchitis/ OR exp Emphysema/ OR OR OR OR OR (obstructive AND lung AND disease).mp.) AND (exp Respiratory Failure/ OR (respiratory AND failure).mp.)

Cochrane: 'doxapram'

Search Outcome

36 papers were identified by searching the databases, of which 2 were relevant. 2 case reports were excluded due to the low level of evidence. A further trial was found through searching the reference lists.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Sheppard JD.
14 patients with alveolar hypoventilation syndrome and respiratory failure with elevated PaCO2; 13 with chronic obstructive pulmonary disease and one with obesity hypoventilatative syndrome. All patients selected had received standard therapy and were alert and responsive prior to being commenced on doxapram.Retrospective case series, level IV.Treatment failure as defined by CNS depression requiring mechanical ventilation.4/14 (29%) patients treated with doxapram required mechanical ventilation. 10/14 (71%) did not require mechanical ventilation and of these 9 survived until discharge and 1 died of a cardiac arrythmia after they had been successfully treated but prior to discharge.Retrospective case series with multiple limitations. No blinding; no control group; small number of patients. Heterogeneous group due to broad inclusion criteria (one patient with pulmonary hypoventilation syndrome; 2 pneumothoraces, 1 overdose).
Moser KM et al.
78 patients aged between 21 and 78 (mean 56.2) with chronic obstructive pulmonary disease and acute respiratory failure, defined by a PaO2 < 50mmHg and PaCO2 > 50 mmHg. The patients were randomised to doxapram vs placebo. The double-blind study period lasted 2 hours with arterial blood gases taken at 0, 30, 60, 90 and 180 minutes.Randomised, placebo controlled double blind trial (level 1b).Treatment failure as defined by an increase in PaCO2 of > 10 mmHg or a pH decline of more than 0.06 with an initial pH of > 7.35 or 0.04 with an initial pH of < 7.35 or if invasive ventilation deemed necessary. When treatment failure defined by both PaCO2 and pH, 36.8 % of the placebo group and 17.5 % of the doxapram group were failures (p < 0.05). Treatment failure defined on PaCO2 alone, 28.9 % of the placebo group were failures compared with 5 % of the doxapram group (p < 0.05). On pH criteria alone, 31.5 % of the placebo group were treatment failures compared to 17.5 % of the doxapram group (p < 0.05). One patient in the doxapram group was a treatment failure due to requiring mechanical ventilation before the end of the 2 hour study period, no comment was made on whether any of the placebo group were treatment failures for this reason.Low patient numbers. Study across 8 centers with no protocol in place for the use of adjuncts (anti-biotics, bronchodilators or expectorants), this was acknowledged in the study. No comment on the randomisation process although assessment of the populations by the investigators did not demonstrate a significant difference between placebo and doxapram group. Age range of 21-78 suggests heterogenecity between patients in terms of severity of disease, no comment of this in the study.
Long term outcome of patients in terms of need for mechanical ventilation or death.No correlation in any systematic way with the data obtained before or at the conclusion of the 2 hour period.
Riordan et al
8 patients aged between 45 and 79 (mean 63) admitted to hospital due to an acute exacerbation of chronic bronchitis with persisting respiratory failure as defined by a PaO2 < 55 mmHg and a PaCO2 of > 50mmHg despite initial treatment. Patients admitted to the study were judged to be sufficiently stable to allow temporary withdrawal of all therapy except antibiotics and physiotherapy. Patients were divided into two groups, Group I received placebo for the first 3 hour treatment period and Group II received doxapram for the first 3 hour treatment period. The patients then changed treatment for the next 3 hour period.Controlled, double-blind, crossover trial (Level 2a).Unable to complete protocol due to CO2 narcosis.One patient in Group I was unable to complete the protocol, all patients in Group II were able to complete the protocol.Small study (n=8). No objective protocol dictating who should enter trial, based on subjective opinion of investigators. No comment of randomisation procedure. No comment on any significant difference between groups. No protocol for use of anti-biotics or amount of physiotherapy patients received. Wide variation between admission to hospital and entry to trial (2-8 days, mean 3.7) suggests that not all patients acutely retaining carbon dioxide, some may have been chronic retainers of carbon dioxide who were back to their baseline on entry to the trial. Although acknowledged by the investigators the doxapram used was donated by a pharmaceutical company, creating the potential for bias.
PaCO2 changes.PaCO2 was lower after doxapram in each patient that completed the protocol; the mean difference of 6.8 mmHg was highly significant (P< 0.01).
Greenstone et al
The authors searched the Cochrane Airways Group trials register for relevant trials using the search term ‘doxapram*’. This register is sourced by Medline, CINAHL and EMBASE. Reference lists were searched and diagnostic companies contacted. Only the trial by Moser et al was included for the comparison between doxapram and placebo Cochrane Systematic ReviewFrequency of unsatisfactory response (doxapram vs placebo). Unsatisfactory response defined as a decrease in pH >0.06 or a rise in paCO2 >10 mm Hg (1.3 kPa), or if intubation or tracheostomy was necessary. OR 0.38 (95% CI 0.14 to 1.02), favouring doxapram (not statistically significant)2 reviewers searched for and appraised papers but it is not clear whether they each appraised the papers identified. There is no assessment of interobserver reliability. Subjective criteria for quality review.
Need for intubation and ventilationOR 1.29 (95% CI 0.51 to 3.27), favouring placebo (not statistically significant)


We did not identify any trials that directly answered the question posed. We did identify one placebo-controlled randomised controlled trial, one double-blind crossover controlled trial and one retrospective case series that evaluated the use of doxapram without non-invasive ventilation, although all of these papers were published in the 1970s and had methodological weaknesses. In particular, small numbers precluded a precise evaluation of the impact of doxapram on mortality, which is arguably the most important clinical outcome. The papers identified do, however, provide weak evidence to suggest that doxapram improves the biochemical markers of disease severity. All of the studies identified suggested that the drug was well tolerated. To provide a definitive answer, appropriately powered randomised controlled trials would be necessary to evaluate not only the effect on biochemical markers but also on clinical outcomes. Given that the majority of patients now receive non-invasive ventilation, it is unlikely that an adequately powered randomised controlled trial to compare doxapram with placebo in those unable or unwilling to undergo non-invasive ventilation will be practical. Interestingly, however, the systematic review by Greenstone et al identified that there is currently no evidence to demonstrate that non-invasive ventilation is superior to doxapram, which is perhaps a more pragmatic target for future research

Clinical Bottom Line

In patients who are ineligible for non-invasive or invasive ventilation and who do not respond to initial medical therapy, the available evidence suggests that the use of doxapram should be considered.

Level of Evidence

Level 2 - Studies considered were neither 1 or 3.


  1. Sheppard JD. Experience with the use of doxapram in the treatment of respiratory failure. Journal of The National Medical Association 1972 July; 64 (4); 342-344.
  2. Moser KM, Luchsinger PC, Adamson JS et al. Respiratory stimulation with intravenous doxapram in respiratory failure. A double-blind co-operative study. The New England Journal of Medicine. March 1, 1973. Volume 288 (9). 427-431.
  3. Riordan JF, Sillet RW and McNicol MW. A Controlled Trial of Doxapram in Acute Respiratory Failure. British Journal of Diseases of the Chest. 1975; 69: 57-62.
  4. Greestone M and Lasserson TJ. Doxapram for ventilatory failure due to exacerbations of chronic obstructive pulmonary disease. Cochrane Database of Systematic Reviews. 2002 , Issue 3.