Cutler R B, Fishbain D A, Lu Y, Rosomoff R S, Rosomoff H L
Department of Neurological Surgery, University of Miami, School of Medicine, FL 33139.
Clin J Pain. 1994 Mar;10(1):10-7. doi: 10.1097/00002508-199403000-00003.
Geriatric chronic pain patients (age 65 and over) form an increasing percentage of the pain center treatment population. It is therefore important to be able to predict pain center treatment success or failure for these patients; this is the first study to address this concern.
Chronic pain patients rated themselves at pain center admission and discharge on 43 rating scales for the areas of pain, functional status, behavioral variables, and other pain center modification categories. The 43 scores at admission were used as potential predictors, while the 43 change scores (from admission to discharge) were the outcome measures to be predicted. Additional possible predictors were 16 other variables that are considered prognostic of treatment outcome, including age, number of surgeries, and prior occupational level. The statistical analysis consisted of a five-step procedure: (a) mathematical techniques were used to remove redundant outcome measures; (b) each of the remaining outcome variables was correlated with the full set of predictor variables; (c) regression techniques were used to predict the outcome variables; (d) these outcome variables were combined into independent factors using factor analysis; and (e) regression techniques were used to predict the factors.
The variable-reduction technique was successful in removing 26 of the 43 outcome variables. Factor analysis of change scores of the remaining variables resulted in four factors, which were identified as change in activity, change in pain and behavior, change in constant pain, and change in attitude to pain center goals. The analysis showed that the best predictor of a variable's change score was the initial level of that variable. Regression analysis, using all variables as predictors except initial level, found a number of statistically significant predictors. However, no predictor variable, alone or in combination, was able to account for > 30% of the variance of any outcome measure.
These results indicate that we cannot as yet predict geriatric pain center treatment outcome. Potential reasons for these results are discussed.
老年慢性疼痛患者（65岁及以上）在疼痛中心治疗人群中所占比例日益增加。因此，能够预测这些患者在疼痛中心的治疗成败至关重要；这是第一项针对这一问题的研究。
慢性疼痛患者在疼痛中心入院和出院时，就疼痛、功能状态、行为变量及其他疼痛中心调整类别等领域，通过43个评定量表进行自我评定。入院时的43个评分用作潜在预测指标，而43个变化评分（从入院到出院）则是要预测的结果指标。其他可能的预测指标包括另外16个被认为可预测治疗结果的变量，如年龄、手术次数和既往职业水平。统计分析包括五个步骤：(a) 使用数学技术去除冗余的结果指标；(b) 将每个剩余的结果变量与全套预测变量进行相关性分析；(c) 使用回归技术预测结果变量；(d) 通过因子分析将这些结果变量合并为独立因素；(e) 使用回归技术预测这些因素。
变量缩减技术成功去除了43个结果变量中的26个。对剩余变量的变化评分进行因子分析，得出四个因素，分别被确定为活动变化、疼痛和行为变化、持续疼痛变化以及对疼痛中心目标的态度变化。分析表明，变量变化评分的最佳预测指标是该变量的初始水平。在回归分析中，除初始水平外，将所有变量用作预测指标，发现了一些具有统计学意义的预测指标。然而，没有任何一个预测变量单独或联合起来能够解释任何结果指标超过30%的方差。
这些结果表明，我们目前还无法预测老年疼痛中心的治疗结果。文中讨论了导致这些结果的潜在原因。
