SEARCH · PubMed Health

Results for “multivariate analysis”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 235 records · Page 13Linked to original sources

PubMed2000

Shelf-life prediction of processed milk by solid-phase microextraction, mass spectrometry, and multivariate analysis.

A technique based on solid-phase microextraction, mass spectrometry, and multivariate analysis (SPME-MS-MVA) was used to predict the shelf life of pasteurized and homogenized reduced-fat milk and whole-fat chocolate milk sampled over a 7 month period. Using SPME-MS-MVA, which is essentially a mass spectrometry-based electronic-nose instrument, volatile bacterial metabolites were extracted from milk with SPME (Carboxen-PDMS) and injected into a GC capillary column at elevated temperature. Mass fragmentation profiles from the unresolved milk volatile components were normalized to the intensity of a chlorobenzene internal standard mass peak (m/z 112) and subjected to MVA. Prediction models based on partial least-squares regression of mass intensity lists were able to predict the shelf life of samples to approximately +/-1 day, with correlation coefficients greater than 0.98 for the two types of milk samples. Using principal component analysis techniques, the procedure was also useful for classifying samples that were rendered unpalatable by nonmicrobial sources (contamination by copper and sanitizer) as well as by bacteria.

Animals↗

PubMed1992

Multivariate analysis of postoperative complications after esophageal resection.

To determine the contributing factors for eight postoperative complications after esophagectomy through a right thoracoabdominal approach, a multivariate analysis was carried out on preoperative and intraoperative variables in 141 patients with thoracic esophageal cancer. Although postoperative complications occurred in 125 patients, only 7 died of such complications. The multivariate analysis indicated that the retrosternal route was a significant factor predisposing to postoperative atelectasis. Age, preoperative arterial oxygen tension, and volume transfused were significant factors predisposing to postoperative hypoxemia, whereas age, routes other than the intrathoracic route, and volume transfused were significant factors predisposing to prolonged respiratory support. In addition, preoperative total serum bilirubin level and volume transfused were significant factors predisposing to postoperative hyperbilirubinemia; preoperative serum creatinine level was a significant contributing factor for postoperative renal insufficiency; and sex, antesternal route, and substituted colon were significant contributing factors for anastomotic leakage. There were no significant factors predisposing to postoperative pneumonia and liver dysfunction. These significant factors should be taken into consideration not only during perioperative management but also when choosing the operative procedures and extending the surgical indication for esophagectomy through a right thoracoabdominal approach.

Age Factors↗

PubMed1993

Multivariate analysis of prognostic factors in patients with non HIV-related primary cerebral lymphoma. A proposal for a prognostic scoring.

Between 1982 and 1991, 41 patients were treated for non HIV related primary cerebral lymphoma (PCL) in our institute. The purpose of this study was to perform a multivariate analysis of prognostic factors for survival in these patients. The presence of a CSF protein level over 0.6 g l-1 at diagnosis was found to be the most significant unfavourable prognostic factor in univariate analysis and had not previously been reported. Among the five significant prognostic factors at diagnosis, (age over 60 years, performance status--ECOG scale--over 2, memory dysfunction, non hemispheric tumour site, CSF protein level over 0.6 g l-1 at the diagnosis), three independent factors were identified in multivariate analysis: (1) CSF protein level (P = 0.007; RR = 4.7); (2) PS > 2 (P = 0.04, RR = 2.65); (3) age over 60 (P = 0.08; RR = 2.43). Using the regression coefficient of these three parameters, we determined a prognostic index which allowed us to distinguish three risk groups whose theoretical median survival is 4, 20 and 54, months respectively in patients with non HIV related PCL. These results indicate that PCL is an heterogeneous disease in terms of the prognostic in which three subgroups with discriminant survival can be identified.

Adolescent↗

PubMedDate not supplied

Papillary thyroid carcinoma: a multivariate analysis of prognostic factors including an evaluation of the p-TNM staging system.

OBJECTIVE: To analyse the prognostic factors in papillary thyroid carcinoma, and in particular to evaluate the accuracy of the pathological tumour, nodes, metastases (p-TNM) staging. DESIGN: Retrospective univariate and multivariate analysis. SETTING: University hospital in Norway. SUBJECTS: 167 patients who were operated on for papillary thyroid carcinoma between 1971 and 1985. Main outcome measures--Death of papillary thyroid carcinoma, and length of recurrence free survival. RESULTS: Male sex, increasing age, larger tumours, and spread of growth beyond the thyroid all independently increased the risk of dying of papillary thyroid carcinoma, whereas the period of recurrence free survival was influenced only by the presence of regional metastases and the patient's age. The age related p-TNM staging is suitable for predicting the likelihood of death, but is less accurate in the prediction of recurrence free survival. The age of 45 years is too low to be useful in predicting survival, especially in women. CONCLUSION: The identification of sex in the multivariate analysis as a strong independent predictor of death of papillary thyroid cancer suggests that the prognostic value of the age related p-TNM staging system could be improved if sex was adjusted for, and if a different age was used for men and women.

Adult↗

PubMed1997

Invasive carcinoma in colorectal adenomas: multivariate analysis of patient and adenoma characteristics.

BACKGROUND AND STUDY AIMS: The risk of invasive carcinoma developing in colorectal adenomas is influenced by a number of characteristics, relating both to the patients and to the adenomas, and by the composition of the sample analyzed. The aim of the present study was use a multivariate analysis to investigate the risk of invasive carcinoma in endoscopically and surgically removed adenomas. PATIENTS AND METHODS: Between 1978 and 1993, more than 20,000 polyps were prospectively documented at the Erlangen Registry of Colorectal Polyps. A multivariate analysis of 11,188 adenomas detected at the first total colonoscopy was carried out in order to investigate the risks associated with size and site--both of which can be assessed by endoscopic inspection alone- and the extent to which these may be modified by other patient and adenoma characteristics, with an influence on the risk of invasive carcinoma in colorectal adenomas. RESULTS: The size of the adenoma proved to be the most important influencing factor. Invasive carcinoma was never found in 5027 small adenomas (< or = 5 mm). Adenomas in the right colon had a lower risk than those in the left colon or rectum. But with increasing adenoma size, the malignancy rate showed a right-sided shift, with a significant interactive effect of size and right-sided location. However, the risk determined by the size and site of the adenoma was significantly modified by a number of patient and adenoma characteristics, including histological type, presence of multiple adenomas, and patient age and sex. CONCLUSIONS: The risk of invasive carcinoma in colorectal adenomas can only be adequately described by a complex model of the interactive effects of patient and adenoma characteristics on the main factors of size and site.

Adenoma↗

PubMed1995

Polymorphisms in alcohol metabolizing enzyme genes and alcoholic cirrhosis in Japanese patients: a multivariate analysis.

Alcohol dehydrogenase (ADH), aldehyde dehydrogenase (ALDH), and P450IIE1 are the primary enzymes that catalyze the conversion of ethanol to acetaldehyde and then to acetate. Genetic polymorphisms have been reported in ADH2, ADH3, ALDH2, and the 5'-flanking region of P450IIEI. In this study, we used multivariate analysis to determine which genetic polymorphisms in alcohol metabolizing enzymes were independently associated with the development of alcoholic cirrhosis. Thirty-four noncirrhotic alcoholic patients, including 27 with fatty liver and 7 with nonspecific changes, and 46 patients with alcoholic liver cirrhosis were studied. Restriction fragment length polymorphisms (RFLPs) in the ADH2 and P450IIE1 genes were detected by digestion of polymerase chain reaction (PCR)-amplified DNA with MaeIII and RsaI, respectively. In the ALDH2 gene, RFLPs were detected by differences in the MboII site after PCR amplification. By multivariate analysis of four significant factors including total alcohol intake, ADH, ALDH, and P450IIE1 using the multiple logistic regression model, genotype ADH2(2)/ADH2(2) (P = .029) and genotype c1/c1 of P450IIE1 (P = .013) were found to be independently associated with alcoholic cirrhosis. The odds ratios for ADH2(2)/ADH2(2) genotype and the type A genotype of P450IIE1 (c1/c1) were 4.600 and 4.006, respectively. These results suggest that ADH2 and P450IIE1 gene polymorphisms may be independently associated with the development of alcoholic liver cirrhosis in Japan.

Adult↗

PubMed1995

Prognostic value of echocardiography compared to other clinical findings. Multivariate analysis based on long-term survival in 456 patients.

The prognostic significance of conventional clinical and echocardiographic data in patients referred to echocardiography was retrospectively analyzed. 456 patients (206 females and 250 males) were studied in the department of cardiology in a district hospital. Survival after 3 years was 64%. By multivariate analysis five factors contained independent, significant, prognostic information (hazard ratios for death are given in parentheses): left ventricular wall motion score index (WMI) < or = 1.2 by echocardiography (2.5), status as inpatient (2.1), age > 65 years (1.7), clinical heart failure (1.9) and atrial fibrillation (1.5). A stepwise multivariate analysis was performed by entering variables into a model initially forced to contain information on age, hospitalization status, treatment of heart failure and heart rhythm. In this analysis, a poor WMI (< or = 1.2) and a dilated right ventricle contained further independent prognostic information. In conclusion, among conventional clinical and echocardiographic data, WMI was the most powerful predictor of long-term survival, and despite knowledge of major clinical features echocardiography provided further prognostic information.

Aged↗

PubMed1991

Radiotherapy in T1 laryngeal cancer: prognostic factors for locoregional control and survival, uni- and multivariate analysis.

From 1975 through 1985, 194 patients with T1 glottic, 37 patients with T1 supraglottic, and 3 patients with T1 subglottic cancer were treated with radiotherapy. Local control and ultimate locoregional control (after salvage surgery) was 91% and 97% for T1 glottic, 84% and 81% for T1 supraglottic, and 2/3 and 3/3, respectively for subglottic tumors. In uni- and multivariate analysis local control for glottic tumors was associated with extension of the tumor on the vocal cord (entire length of vocal cord vs others, p = 0.01) and continuation of smoking after therapy (yes/no, p = 0.03). No prognostic factor for local control was found in supraglottic tumors. However, regional control and survival were impaired by N stage (N0 vs N+, p less than 0.0005), local recurrence (yes/no, p less than 0.0005), and extension of the tumor (one supraglottic subsite vs more than one, p less than 0.05). Mild late complications were seen in 13% of patients without salvage therapy. Following univariate analysis, field size, fraction size (greater than 2 Gy), maximum tumor dose (greater than 70 Gy), age, post-treatment biopsy, and tumor site were associated with complication rate. Following multivariate analysis, site, fraction size, maximum tumor dose, and continuation of smoking after therapy were independent prognostic factors for mild late complications (mostly arytenoid edema).

Adult↗

PubMed1993

Multivariate analysis of the relationship between survival and the microstage of primary melanoma by Clark level and Breslow thickness.

BACKGROUND: The American Joint Committee on Cancer (AJCC) uses both Breslow thickness and Clark level in its staging system for malignant melanoma. Stage I corresponds to Breslow thicknesses less than 1.5 mm and Clark levels II and III. Stage II corresponds to Breslow thicknesses of at least 1.5 mm and Clark levels IV and V. However, most investigators have found Clark level to be of no prognostic significance once Breslow thickness has been taken into consideration by multivariate analysis. METHODS: The authors examined the prognostic significance of Clark level by studying patients in the large database of the John Wayne Cancer Institute. Among 5575 patients with melanoma seen during the past 20 years, complete data on microstaging by both Clark and Breslow methods were available for 3323 patients. The 5-year survival rates were as follows: Clark II, 95%; III, 81%; IV, 68%; V, 47%. The Breslow thicknesses were as follows: < 0.75 mm, 95%; 0.75-1.49 mm, 85%; 1.5-3.99 mm, 66%; > or = 4.0 mm, 46%. RESULTS: By univariate analysis, both Clark level and Breslow thickness were highly significant prognostic indicators (P < 0.0001). By multivariate analysis, Breslow thickness remained significant (P < 0.0001). However, even when Breslow thickness was included in the model, Clark level also remained highly significant (P < 0.0015). CONCLUSIONS: Decisions regarding therapy for patients with clinical Stage I melanoma should consider both Clark level and Breslow thickness of the primary lesion. When there is a discordance between the two methods of microstaging, the AJCC stage should be amended to reflect the least favorable of the two prognostic indicators.

Humans↗

PubMed1993

[Laser therapy of villous rectocolonic tumors. Evaluation of results by multivariate analysis].

One hundred and twenty-four patients (mean age: 70 years) with a villous tumour of the rectum (n = 98) or the colon (n = 26) were treated by laser therapy from 1985 to 1991. Initial complete eradication was obtained in 93.5% of cases with a median time of 16 months. Using multivariate analysis, the tumor size (> 4 cm) only was predictive of eradication (P < 0.001). At 6 and 12 months, the actuarial eradication rates were 91 and 98% for small tumours versus 45 and 74% for larger tumours respectively. The actuarial recurrence rates were 10.5, 23, and 37% at 6, 12 and 24 months, and no longer increased after 3 years. No predictive factor of recurrence was isolated by multivariate analysis. Seven complications (5.6%) and 4 malignant transformations (3.4%) occurred. This study confirms the efficiency and low morbidity of laser therapy, especially in patients with small tumors, less than 4 cm in size.

Adenoma, Villous↗

PubMed1999

Prognostic factors in fetal hydronephrosis: a multivariate analysis.

With the increasing use of obstetric echography fetal hydronephrosis has been reported more frequently. The purpose of this study was to identify prognostic factors associated with adverse outcome, such as renal failure and death, in fetal hydronephrosis. One hundred and forty-eight children with fetal hydronephrosis were admitted, submitted to a systematic protocol, and prospectively followed. Prognostic factors associated with fetal echography and clinical and laboratory findings on admission were studied. The median follow-up was 39 months. The analysis was conducted in two steps. In a univariate analysis, variables associated with adverse outcome were identified by the Kaplan-Meier method. The variables that were significantly associated with adverse outcome were then included in a multivariate analysis. This analysis, using the multivariate Cox's model, was performed to identify variables that were independently associated with a worse prognosis. Only variables that remained independently associated with adverse outcome were included in the final model. After final adjustment by Cox's multivariate model, three variables were identified as independent predictors of adverse outcome: oligohydramnios, prematurity, and glomerular filtration rate lower than 20 ml/min. Thus, in the presence of oligohydramnios, prematurity, and abnormal renal function, the medical team must plan appropriate follow-up for infants at health centers prepared to investigate and treat uropathies in newborns.

Algorithms↗

PubMed2006

Medullary thyroid carcinoma: multivariate analysis of prognostic factors influencing survival.

BACKGROUND: Medullary thyroid carcinoma (MTC) is a rare development of thyroid cancer with a no negligible mortality rate. Our aim was to determine factors that predict outcome in patients with MTC. METHODS: We reviewed the records of all patients with MTC (n=56) who underwent treatment at our institution between January 1990 and December 2000. Univariate and multivariate analysis of clinicopathologic predictors of MTC outcome were performed to identify subsets of patients with different probabilities in terms of overall survival, local recurrence, and distant metastases. RESULTS: Multivariate analysis demonstrated that a statistically significant decrease in overall survival is associated with T4b tumours (p=0.06), the presence of distant metastases at the time of presentation (p=0.033), lymphatic invasion (p=0.099), and postoperative treatment (p=0.045). CONCLUSIONS: The analysis of survival curves of patients with MTC shows that the occurrence of locoregional and distant metastases occurs preferentially within the first 5 years, which identifies this as a crucial period for follow-up. In this series of patients with MTC, the tumours classified as T4b, metastases at presentation, the presence of lymphovascular invasion, and postoperative treatment were the most important prognostic features. At present, there is no available beneficial adjuvant therapy. However, as the development of molecular therapy progresses, it should be tested in clinical trials with the purpose of achievement of novel targeted therapies for selected MTC patients with risk factors.

Adult↗

PubMed2005

Predictions of outcomes of renal stones after extracorporeal shock wave lithotripsy from stone characteristics determined by unenhanced helical computed tomography: a multivariate analysis.

The aim of our study is to analyze the relationships between the characteristics of renal stones determined by unenhanced helical computed tomography (UHCT) and their outcomes after extracorporeal shock wave lithotripsy (ESWL) as well as to predict ESWL outcomes of renal stones by their UHCT characteristics with the use of multivariate analysis. During a 7-month period, 80 adult patients with renal stones underwent ESWL as well as UHCT both before and 3 months after ESWL. Of the 80 patients, 42 patients were classified as ESWL successes and 38 as ESWL failures based on their post-ESWL UHCT findings. For pre-ESWL UHCT, a stone number of more than 2 (P=0.0236), a maximal stone size of greater than 12 mm (P<0.0001), a stone burden of more than 700 mm3 (P<0.0001), a maximal stone density of more than 900 HU (P=0.0008) and nonround/oval stones (P=0.0007) were associated with ESWL failure outcomes. Multivariate analysis demonstrated that a stone burden of more than 700 mm3 (P=0.0003), the presence of nonround/oval stones (P=0.0072) and a maximal stone density of more than 900 HU (P=0.0430) were statistically significant predictors of a failure outcome for ESWL. Thus, the analysis of stone characteristics of renal stones by UHCT is helpful in selecting appropriate patients undergoing ESWL for favorable outcomes and reduces the overall costs of the treatment of renal stones.

Adult↗

PubMed1996

High-dose chemotherapy as salvage treatment in germ cell tumors: a multivariate analysis of prognostic variables.

PURPOSE: To identify prognostic variables for response and survival in male patients with relapsed or refractory germ cell tumors treated with high-dose chemotherapy (HDCT) and hematopoietic progenitor cell support. PATIENTS AND METHODS: Three hundred ten patients treated with HDCT at four centers in the United States and Europe were retrospectively evaluated. Univariate and multivariate analysis of patient, disease, and treatment characteristics were used for comparisons of response rates and failure-free survival (FFS). RESULTS: The actuarial FFS rate was 32% at 1, 30% at 2, and 29% at 3 years. Multivariate analysis identified progressive disease before HDCT, mediastinal nonseminomatous primary tumor, refractory or absolute refractory disease to conventional-dose cisplatin, and human chorionic gonadotropin (HCG) levels greater than 1,000 U/L before HDCT as independent adverse prognostic variables for FFS after HDCT. These variables were used to identify patients with good, intermediate, and poor prognoses. In the good-risk category, the predicted FFS rate at 2 years was 51%, compared with 27% and 5% in the intermediate-risk and poor-risk categories (P < .001). The increased risk for treatment failure was due to both a significantly lower rate of favorable responses and a significantly higher rate of relapses. Within the prognostic categories, the particular HDCT regimen or higher dosages of carboplatin or etoposide did not have a significant influence on treatment outcome. CONCLUSION: Prognostic variables for treatment response after HDCT can be identified. The proposed prognostic model might help to optimize the use of HDCT in germ cell tumors and warrants validation in future trials.

Antineoplastic Combined Chemotherapy Protocols↗

PubMed1997

Squamous cell carcinoma of the penis: multivariate analysis of prognostic factors and natural history in monocentric study with a conservative policy.

BACKGROUND: Penile carcinoma is uncommon in Western countries. Here we report on a large series of patients with squamous cell carcinoma (SCC) of the penis, describing prognostic factors, survival and therapeutic results. PATIENTS AND METHODS: From 1973 to 1993, 102 patients with invasive SCC of the penis were treated at the Institut Gustave-Roussy. Precancerous lesions and conditions associated with penis cancer were analyzed retrospectively. Survival curves were estimated by the Kaplan-Meier method, and groups were compared for outcome by the log rank test for univariate comparisons and by Cox's proportional hazards model for multivariate analysis. RESULTS: The median age at onset was 58 years. Sixty-nine patients presented with Jackson's stage I disease, 17 with stage II and 15 with stage III. The interval between the manifestation of symptoms and the diagnosis was more than a year in 13.7% of cases. Precancerous lesions were found in 17 (16.6%) patients, and a history of phimosis was noted in 25 (24.5%). In situ and invasive carcinoma were observed together in 17 (16.6%) cases and dysplasia was associated with invasive carcinoma in eight (7.8%) further cases. Conservative treatment was administered whenever feasible. Interstitial brachytherapy was performed alone or associated with limited surgery (local excision or circumcision) in 72 (70.6%) patients. Of the 28 patients with a local relapse, nine have died of their neoplasms (32%) compared to 21 of 28 patients with lymph node relapse (75%). The median follow-up was 111 months. Disease-free survival, disease-specific survival and overall survival were, respectively, 56%, 72% and 63% at five years and 42%, 66% and 50% at 10 years. Age (P = 0.01), the N status (P < 0.00001) or palpable nodes (P < 0.0038), corpus involvement (P = 0.006) and a verrucous histology (P = 0.038) had significant prognostic relevance for survival in the univariate analysis whereas the performance status, T status and Broders' grade did not. In the multivariate analysis only two parameters, involvement of the corpus (P < 0.0001) and palpable nodes (P = 0.009), were singled out as being independent variables influencing survival. A subgroup of nine patients with verrucous histologies were distinguished by their freedom from node involvement. These patients had an excellent prognosis: all are alive and disease-free. Penile integrity was preserved during follow-up in 54 patients (52.9%), 31 of whom are still alive. Of 72 patients treated by a conservative approach including brachytherapy, long-term penile integrity was maintained in 49 (68%). CONCLUSION: Corpus involvement and clinically palpable nodes are highly statistically significant independent factors influencing overall survival. Node relapses remain a major cause of death. Thus, better management of lymph nodes is essential for improving survival even when conservative therapy is used to treat the primary.

Adult↗

PubMed1998

Prognostic value of DNA ploidy patterns of colorectal adenocarcinoma: univariate and multivariate analysis.

BACKGROUND/AIM: The prognostic significance of DNA ploidy patterns of colorectal cancer has not yet been settled. The present study was designed to determine the prognostic value of DNA ploidy patterns for colorectal adenocarcinomas after curative resection. METHODS: DNA ploidy patterns of 140 colorectal adenocarcinomas were determined by DNA flow cytometry, and the prognostic significance of DNA ploidy patterns was evaluated by univariate as well as multivariate analysis. RESULTS: DNA ploidy patterns were diploid in 75 (53.6%) and aneuploid in 65 patients (46.4%). DNA ploidy patterns did not correlate with any of conventional prognostic variables. Univariate analysis disclosed that Dukes B2, C1, and C2 stages of the disease (p < 0.01), positive nodal metastases (p < 0.01), invasion through the intestinal wall (p < 0.01), and poor tumor differentiation (p < 0.05) were associated with worsened survival, but no correlation was found between DNA patterns and survival of patients. Multivariate analysis disclosed that tumor penetration through the bowel wall was associated with poorer survival of patients but the DNA ploidy pattern had no prognostic significance. CONCLUSIONS: A significant prognostic variable for patients after curative resection of colorectal adenocarcinoma was penetration of tumor through the bowel wall but not DNA ploidy patterns.

Adenocarcinoma↗

PubMedDate not supplied

Preoperative values of CA 15-3 and CEA as prognostic factors in breast cancer: a multivariate analysis.

The role of circulating tumor markers in providing prognostic information has not been widely studied. In the current study, serum levels of the carbohydrate antigen 15-3 (CA 15-3) and carcinoembryonic antigen (CEA) were determined preoperatively in 364 breast cancer patients with no clinical signs of metastasis. The prognostic relevance of these markers for recurrence (175/364) and death of disease (104/175) was determined by Cox multivariate analysis, including the comparison with classical prognostic factors. High levels of both tumor markers were associated with aneuploid tumors with high S-phase fraction and high ornithine decarboxylase activity. CA 15-3 was highly associated with the number of positive lymph nodes and peritumoral lymphatic or blood vessel invasion. No significant associations were found between CEA or CA 15-3 levels and histologic grade, necrosis and steroid receptor status. In univariate analysis, preoperative values, using optimum cutoff values of CA 15-3 (40 U/ml) and CEA (6 ng/ml), were statistically significant for relapse-free survival and overall survival. In multivariate analysis, only node status, DNA ploidy and ornithine decarboxylase activity were independent predictors for relapse-free survival; the estrogen receptor status was a predictor of overall survival. In node-negative patients, ornithine decarboxylase activity was the only factor selected for relapse-free survival. In node-positive patients, the number of lymph nodes and DNA ploidy were the only variables selected for relapse-free survival or overall survival. Estrogen receptor and ornithine decarboxylase activity were excluded for relapse-free survival, but were significant prognostic factors for overall survival.

Adult↗

PubMed1994

Postoperative bone re-growth in lumbar spinal stenosis. A multivariate analysis of 48 patients.

STUDY DESIGN: Forty-eight patients who underwent posterior decompressive surgery for lumbar spinal stenosis were studied. The effect of bone re-growth after posterior decompression in lumbar spinal stenosis on clinical outcome and the factors promoting the bone re-growth were investigated with a multivariate analysis model. METHODS: Bone re-growth at the sites operated upon was evaluated as a percentage of re-growth of the original laminectomy site based upon plain radiographs. The degree of bone re-growth was classified into four groups: 1) no significant re-growth (Group I, bone regrowth rate 10% or less); mild (Group II, 11% to 40%), moderate (Group III, 41% to 70%), and marked (Group IV, 71% to 100%). The average follow-up period was 4.5 years (range, 2 to 7 years). RESULTS: No significant bone re-growth was shown in 6% of the patients. Mild re-growth was shown in 50%, moderate re-growth in 29%, and marked re-growth in 15%. A multivariate analysis indicated that a total block in the preoperative myelogram, a follow-up period of more than 5 years, decompression at more than three spinal levels, and age under 60 years were associated with moderate or marked bone re-growth. Spinal instability accelerated the bone re-growth mainly in the mid and later follow-up intervals. Spinal levels adjacent to a fusion showed more bone re-growth. Patients with moderate and marked bone re-growth had poorer clinical outcomes than those with no significant and mild bone re-growth. CONCLUSION: Bone re-growth in a surgical defect will occur in most patients after posterior decompression. Moderate and marked postoperative bone re-growth are possibly related to recurrence of neurologic symptoms in the middle of and later on in follow-up periods.

Aged↗