To the Editor:
I read with interest the letter from Gates et al.  concerning sensory loss of the distal phalanx associated with pulse oximeter probe position and would like to report two similar cases.
Case 1. A 40-yr-old woman without significant medical history underwent a 6-h endoscopic sinus surgery under general anesthesia. A Hewlett-Packard reusable oximeter probe M1190A was placed on her right index finger and both arms were tucked at her side. (The Hewlett-Packard oximeter probe M1190A is a rubber-like, nonpadded probe that fits snugly over the middle and distal phalanges.) There was no intraoperative hypotension, and the oximeter gave a reading of 100% during the entire case. In the recovery room the patient noted numbness of the entire middle and distal phalanges of the right index finger. The patient was seen by a neurologist who confirmed the pure sensory loss. No medical basis for the sensory loss was found. The patient was contacted monthly and noted gradual return of sensation, proximally to distally, over a 6-month period.
Case 2. A 21-yr-old woman without any medical problems underwent a 5-hour LeFort I maxillary osteotomy with bone grafting under general anesthesia. A Hewlett-Packard reusable oximeter probe M1190A was placed on her right index finger and both arms were tucked at her side. There was no intraoperative hypotension, and the oximeter gave readings of 99%-100% during the entire case. In the recovery room, the patient complained of numbness of the distal phalanx of her right index finger. This was confirmed on examination in the postanesthesia care unit and on discharge to the floor several hours later. However, the numbness had resolved by the next morning.
The similarities in the above cases lead us to report the occurrences to the Food and Drug Administration Medical Products Reporting Program "Med Watch" in March 1994 and to change somewhat our use of pulse oximeters. First, we no longer use the Hewlett-Packard reusable probe M1190A for procedures greater than 2-3 h in duration or in cases where the patients' arms will be tucked at their side. We use instead the Hewlett-Packard M1900B probe, which utilizes the disposable Nellcor Oxisensor II D-28. We now preferentially place the oximeter probe on the ring finger because we believe that loss of sensation on this digit, should it occur, would be less bothersome for the patient. Such placement has been previously recommended to decrease the occurrence of corneal abrasions . We also take special care when tucking patients' hands that the fingers don't rest under their thigh. (It should be noted that the Hewlett-Packard User's Guide does suggest checking the probe site every 2 h for pressure sores but makes no mention of the potential for neural injury. We were unaware of this suggestion at the time of these cases.)
Unlike in the case reported by Gates et al. , the oximeter functioned normally during both cases, indicating that significant neural damage can occur without the loss of pulse oximetry. The nerve damage in our case was likely due to a combination of neural compression and decreased digital blood flow. That neural injury can occur as described is not surprising as both Lawson et al.  and Greenblott et al.  have shown that pulse oximeters will continue to function with limb blood flow reduced to as little as 8.6% and 25% of baseline limb blood flow.
Pulse oximeters are marvelous devices that likely have significantly increased anesthetic safety; however, their use is not risk free and thought should be used when they are applied.
Patrick J. Donahue, MD
Susan Emery, CRNA, MS
Department of Anesthesia, Deaconess Hospital, Boston, MA 02215
1. Gates RE, Kinsella SB, Moorthy SS. Sensory loss of the distal phalanx and pulse oximeter probe [letter]. Anesth Analg 80:855.
2. Brock-Utne JG, Botz G, Jaffe RA. Perioperative corneal abrasions [letter]. Anesthesiology 1992;77:221.
3. Lawson D, Norley I, Korbon G, et al. Blood flow limits and pulse oximeter signal detection. Anesthesiology 1987;67:599-603.
4. Greenblott GB, Gerschultz S, Tremper KK. Blood flow limits and signal detection comparing five different models of pulse oximeters [letter]. Anesthesiology 1989;70:367-8.