Secondary Logo

Journal Logo


Breath interval as a continuous measure of opioid effects of intravenous fentanyl and alfentanil

Smart, J. A.; Pallett, E. J.; Duthie, David J. R.

Author Information
European Journal of Anaesthesiology: June 2003 - Volume 20 - Issue 6 - p 498-500


The analgesic effects of μ-agonists have never been separated from their depressant effect on spontaneous ventilation. Breath intervals have been shown to be a reproducible method of measuring the time-course of opioid effect in anaesthetized patients breathing spontaneously [1,2]. After a single dose of intravenous (i.v.) opioid, a plot of breath interval (observed – baseline) against time displays a curve very similar to the concentration-time curve after a drug that requires absorption to the circulation (Fig. 1). There is a rise to peak effect because of delays in the uptake of opioid to the brain and the time taken to elicit an effect at the effect site. The duration of the altered breath interval observed after a patient has received an opioid corresponds to the duration of dynamic opioid effect. Relief from pain cannot be measured during anaesthesia, has considerable between-patient variation and can be recorded only intermittently. Breath intervals provide a continuous measure of opioid dynamic effect during anaesthesia in patients breathing spontaneously. We sought to discover if measurement of breath intervals provided a sufficiently sensitive method to distinguish the different duration of action of fentanyl and alfentanil after single i.v. doses.

Figure 1.
Figure 1.:
Serial ‘difference in breath interval’ measurements for two patients given fentanyl (♦) or alfentanil (▴) i.v. at 0 min. Breath intervals are shown after subtraction of baseline values.

Ethics Committee approval and informed written consent were obtained from patients admitted for knee replacement surgery. Patients received temazepam 20 mg before anaesthesia was induced with propofol 2–3 mg kg−1 i.v., and maintained with isoflurane 0.8% and nitrous oxide 67% in oxygen through a laryngeal mask. Sciatic and three-in-one femoral nerve blocks were then performed using a nerve stimulator and bupivacaine 0.375% 20 mL into each nerve.

The analogue output of the carbon dioxide (CO2) concentration from a Capnomac Ultima® (Datex, Helsinki, Finland) was fed to an analogue-to-digital converter sampling at 100 Hz. The time between corresponding points on the down stroke of the CO2-time waveform on successive waves was recorded as the breath interval. Patients were randomized to receive either fentanyl 0.75 μg kg−1 or alfentanil 2.25 μg kg−1 by single injection i.v. at zero time. Breath interval data were plotted against time (Fig. 1). From the latter two-thirds of the elimination phase, the slope of the regression line through the natural logarithm of breath interval – baseline gave the elimination rate constant, k. From this, the dynamic elimination half-life of the opioid effect was calculated using t1/2 = ln(2)/k. The mean effect time [3] of fentanyl in the body was also calculated from the ratio of the area under the first moment curve (AUMC) to the AUC. Both AUC and AUMC were extrapolated to zero by dividing the last datum point by k. The data were analysed using MKMODEL Software [4]. Statistical analysis was by ANOVA. Significance was accepted when P < 0.05.

Thirty-four patients were studied. Sixteen were removed from the study because of inadequate block (n = 8), respiratory depression (6), operation changed (1) and computer failure (1) – leaving nine patients in each group. There were no differences between fentanyl (F) and alfentanil (A) with respect to weight (mean ± SD): F, 66 (±8); A, 77 (±18)kg; height: F, 166 (±8); A, 165 (±11)cm; age: F, 68 (±17) (range 33–86); A, 61 (±19) (33–80) yr; or time from the induction of anaesthesia to the opioid dose i.v.: F, 32 (±14); A, 35 (±15) min. There were four male patients in each group.

Nine patients received fentanyl 0.71 (0.09) μg kg−1 and nine patients alfentanil 2.3 (0.2) μg kg−1. The opioid dynamic effects were significantly different for fentanyl and alfentanil: rate constant of elimination (k): F, 0.0628 (0.0405); A, 0.21 (0.133) min−1, P = 0.009; elimination half-life: F, 15.2 (8.56); A, 5.04 (2.91) min, P = 0.004; mean effect residence time (MRT): F, 23.7 (9.3); A, 7.0 (3.4) min, P < 0.001.

Baseline and peak values for breath intervals and end-tidal CO2 and the time from opioid injection are given in Table 1. There were no differences between the time to peak values of breath interval and CO2 for fentanyl (breath interval – CO2: −0.71 (−2.8 to 1.4) (mean (95% CI)) min, P = 0.48), or alfentanil (breath interval – CO2: −0.79 (−2.2 to 0.6) min, P = 0.25).

Table 1
Table 1:
Baseline and peak breath interval and end-tidal carbon dioxide concentrations.

Breath intervals have proved sufficiently sensitive to distinguish the time-course of dynamic effects of fentanyl and alfentanil. Results obtained show a difference in both the mean residence time of fentanyl and alfentanil at the effect site, and the elimination half-life of the two drugs.

The kinetic and dynamic effects of fentanyl and alfentanil have been modelled from plasma drug concentrations and the EEG [5]. The times to EEG peak effects were 5 min for fentanyl and 1.5 min for alfentanil. These are similar to the times to peak effect by breath interval in this study of 5.3 (1.2) min for fentanyl and 1.9 (1.2) min for alfentanil (Table 1). Modelling kinetic data to obtain the effect site concentrations, the peak effect site concentrations obtained occurred after 1.4 min for alfentanil and 3.6 min for fentanyl after single i.v. injections [6]. The results were similar to the dynamic peak effects measured in this study using breath interval.

The EEG simulation also produced keo, a rate constant of equilibration with the effect site [5]. keo for fentanyl was 0.11 min−1, which was lower than that of alfentanil, 0.63 min−1. The lower keo for fentanyl damps the rise in fentanyl concentrations at the effect site when plasma fentanyl concentrations are at a peak. This may explain why doses of fentanyl in this study that were only one-third of doses of alfentanil produced the same maximum end-tidal CO2 concentrations: fentanyl 7.7 (1.2)%, alfentanil 7.5 (1.2)% (Table 1). In volunteers whose respiratory effects were measured by their response to inhaled CO2 at intervals over 60 min, a dose of alfentanil 20 μg kg−1 was equivalent to fentanyl 2 μg kg−1. The doses of fentanyl 0.75 μg kg−1 and alfentanil 2.25 μg kg−1 in this study were derived empirically during previous clinical practice, being doses which produced a measurable change in breath interval without apnoea, in anaesthetized patients breathing isoflurane and nitrous oxide. The higher keo of alfentanil will produce proportionately higher peak effect site concentrations than fentanyl after a single dose i.v. Limited by apnoea at the peak effect, we were obliged to use a relatively smaller dose of alfentanil. Equianalgesic doses would have produced apnoea in the alfentanil patients or a slight deviation of breath interval only in the fentanyl patients.

There were no differences between the times to peak breath interval and CO2 effect. This suggests both variables were affected by a common stimulus, the opioid, rather than by a change in one being consequent to the effect of the other.

Breath interval to measure opioid effect was successful during surgery only with an effective peripheral nerve block. Patients withdrawn because of respiratory depression had elimination breath interval slopes no different from patients with complete data once spontaneous ventilation resumed. Dynamic effects measured by breath interval had a similar time-course to simulated effect site concentrations and peak opioid effects as reported previously.


1. Smart JA, Pallett EJ, Duthie DJR. Breath interval as a measure of dynamic opioid effect. Br J Anaesth 2000; 84: 735-738.
2. Goodman NW, Vanner RG, Wade JA. Effects of incremental alfentanil and propofol on the breathing of anaesthetised patients. Br J Anaesth 1989; 65: 548-553.
3. Holford N. MKMODEL, 5th edn. Cambridge, UK: Biosoft, 1994.
4. Cheng H, Gillespie WR, Jusko WJ. Mean Residence Time concepts for non-linear pharmacokinetic systems. Biopharmaceut Drug Depos 1994; 15: 627-641.
5. Ebling WF, Lee EN, Stanski DR. Understanding pharmacokinetics and pharmacodynamics through computer simulation: 1. The comparative clinical profiles of fentanyl and alfentanil. Anesthesiology 1990; 72: 650-658.
6. Shafer SL, Varvel JR. Pharmacokinetics, pharmacodynamics and rational opioid selection. Anesthesiology 1991; 74: 53-63.
© 2003 European Society of Anaesthesiology