Predicting the success of a sympathectomy: a prospective study using discriminant function and multiple regression analysis
Walker, P.M.; Johnston, K.W.
Surgery 87(2): 216-221
1980
ISSN/ISBN: 0039-6060 PMID: 7355393 Document Number: 157107
A prospective study of 72 limbs treated with phenol sympatholysis for inoperable peripheral arterial occlusive disease was carried out in order to determine which factor or combination of factors could most accurately be used to predict the success or the failure of the sympathetic interruption. Seven variables were recorded prior to phenol sympatholysis and the patients were followed up to 3 years in order to determine the degree of success or failure of the sympathectomy. The results of discriminant function analysis of these data showed that the most important variables in predicting the outcome were (1) the level of ankle systolic pressure, (2) the presence or absence of a neuropathy, and (3) the extent of the ischemic damage. The correct outcome was predicted in 87% of the cases. Using these three variables, multiple regression analysis was performed to construct a table showing the percentage chance of a successful outcome from a sympathectomy. The results show that a phenol sympathectomy is likely to be successful in the management of a patient with peripheral arterial occlusive disease if there is no evidence of a somatic (and hence autonomic) neuropathy, if the ankle systolic pressure is above 30 mm Hg, and if the tissue damage is not too extensive (i.e., only rest pain, night pain, or digital gangrene is present).