Another opinion

disease also OCCUT. We retrospectively studied lo00 consecutive patients undergoing Caesarean section in a 13-month period recently. Ourgcneralanaesthesiaratcwas3.f./,andour epidural rate was %.So/, No subarachnoid blocks were performed. Although some of the 35 cases done under gmeral anaesthesia could prhaps have baen done Mdcr epidural, we feel that our %.5% epidural anestbeaia rate for Caesarean section approaches an irrujmcibk atinimum rather than the lipre sugecstcd by Davis. Most patients have attended prenatal classes where the benefits of epidural anesthesia are presented. Epidural anaesthesia is well accepted by our Attending Obstet+ cians, nursing staKand resident physicians. Because of their cooperation we have been able to communicate eflectively with patients from diverse backgrounds including some who speak no English. We use a single shot epidural technique. We find that surgery can be generally started within 15 minutes which is approximately the amount of time it takes to start Caesarean section in all but the most urgent cases. In general, inhalational anaesthesia is reserved for patients with fetal distress, maternal hypotension and documented coagulation disorders. Other reasons for not performing an epidural block such as patient refusal, technical difficulties and progressive neurological Department of A t ~ e ~ t h e ~ i ~ l ~ g y , w. sosa St Eamabas Medical Center. Livingstme. New Jersey 07039, USA A. BoDNta

I read with interest the article entitled 'Preoxygenation-how long? ( A~e~t h e s i o 1983; 38: 9 6 102) in which the authors state that 'most anaesthetic machines do not deliver more than 8 litm of oxygen per minute without using the emergency supply'. I am not sure whether this refers to the maximum calibrated delivery or the maximum possible delivery, bearing in mind that the Rowrate of the emergency supply is not calibrated either.
Using a Wright's Respirometer, which has not been independently recalibrated following manufacture, I have tested both the bypass delivery and the maximum attainable delivery from the oxygen Rotameter (by opening the pin valve fully).
The results from 10 standard Boyk's machines show that the mean maximum oxygen delivery is 44.7 litres per minute from the Rotameters (ran@ 41.247.9) whilst the emergency oxygen delivers a mean of only 38.6 litres per minute (range 35.7-42.4).
It may be of interest to note therefore that the Bow is in fact greater from the Rotameter supply than that from the emergency Bush.

PJ.H. V m
The review of intravenous codeine phosphate by Shanahan et al. (Anneslhesin 1983; 38: 40-3) is of considerable interest, but their conclusions cannot he derived from the data of the cases they describe.
The doses of codeine phosphate used in cases one and three. 1.36 mg/kg and 1.1 mg/kg, exceeded the maximum recommended dose in the scale they quote by 81.3% and 48% respectively. The respiratory and cardiovascular depression produced by bolus do-of this order, during nitrous oxide and halothane anaesthesia, is a normal dose-related response to a combination of three depressant drugs; it is somewhat surprising that no attempt was made to reverse the effects of codeine with intravenous naloxone.
The classification of these responses as ad-TC actions to intravenous codeine phosphate is at wiana with the facts presented, and does little to support the important plea for clearer guidelines to the d v of intravenous codeine phosphate for children.

Dublin 12
Fdme of Pealoo IDP veetll.tor . l u p One of us experienced a failure similar to that described by Shribman ( A n~~t h e s i o 1982; 37: 1044) where the alarm attached to the ventilator end of the inspiratory hose did not detect a dixonnexion of the Y-piece and catheter mount.
Further investigation showed that: tbe prrs~Ure switch within the alarm could respond to very short pressure transients (approx 2 m ss duration); p-um transients were produced at the insphtory port of several ventilators when the Y-piece was separated from the catheter mount (these transients were two to three times the magnitude of the steady pressure produced by the resistance of the inspiratory hose to inspiratory flow); some amplihcation of pressure transients occurred along the narrow bone connecting tube and connecting whistle.
Thac factors could combine so that the pressure transients produced by the sudden start of inspiratory %ow would he greater than the switching pressure of tbe alarm for longer than 2 m see. The alarm would therefore not be activated. A solution to this problem, other than resiting the alarm connexion with the consequent problems of condcnsation and contamination.
is to modify the alarm so that it does not respond to prcspurr pulses of very short duration. The manufacturer (Penlon Limited, Abingdon, Oxon, OX14 3PH) can now supply an electronicaUy modified version that will ignore any airway pnssure change that is abow the switching pressure of the alarm (0.75, 1.25 or 2.5 Lpa, depending on model) provided that the duration of the increase in p w u t is less than 0.7 4ec.
Hoamcr, such a modification prevents the use of the alarm when the inspiratory phase was less than 0.7 sea. We arc in the process of evaluating whether this modification is satisfactory in clinical use.

Another opinion
Thank you for allowing us an opportunity to contribute to thc comspondence on the Penlon IDP Ressurc Failure Alarm for intermittent positive prcssure ventilation.
More than basic knowledge is often required to use an alarm system safely in every situation. Several well known ventilators, including the Manley and Cape designs produce a short duration pressure spike in the breathing system at the instant the inflation valve opens. These pressure spikes persist even when very large leaks exist in the breathing system.
The presence of short duration pressure spikes can confuse an alann sensor. The IDP Alarm activates its alarm circuits if it does not receive a sequence of two signals from a preset pressure switch; one as a result of increasing pressure in the breathing system, and the other as a result of a decrease in pressure in the breathing system.
Recently a modification of the IDP Alarm has been produced. A timing circuit has been added so that a short duration spike will be ignored and the alarm will now only be activated by a sequence of signals which are more than 0.6 seconds apart and greater than 1.25 kPa pressure. Any pressure alarm system used with constant pressure generator ventilator may not alarm if a high resistance connexion is used. The complete system, patient included, must be tested in the check procedure to see that the alarm works in the conditions under which it is being uscd. To aid this a sensing light has been provided: if the light does not flash then the alarm will be activated. It should be noted that this type of alarm should not be used with children or rapid rates of ventilation (more than 30/min).
We have had the opportunity to test this device with many different users. We would be interested to know if anyone has had any problems with the new Penlon alann.