Predicting relapse in Crohn's disease: a biopsychosocial model.

BACKGROUND: Crohn's disease (CD) is a chronic relapsing inflammatory bowel disorder. Both biological and psychosocial factors may modulate the illness experience. AIM: The aim of this study was to identify clinical, biological and psychosocial parameters as predictors of clinical relapse in quiescent CD. METHODS: Patients in medically induced remission were followed prospectively for 1 year, or less if they relapsed. Disease characteristics were determined at baseline. Serum cytokines, anti-Saccharomyces cerevisiae antibodies, C-reactive protein (CRP), erythrocyte sedimentation rate and intestinal permeability were measured every 3 months. Psychological distress, perceived stress, minor life stressors and coping strategies were measured monthly. A time-dependent multivariate Cox regression model determined predictors of time to relapse. RESULTS: 101 patients (60 females, 41 males) were recruited. Fourteen withdrew and 37 relapsed. CRP (HR = 1.5 per 10 mg/l, 95% CI 1.1 to 1.9, p = 0.007), fistulising disease (HR = 3.2, 95% CI, 1.1 to 9.4, p = 0.04), colitis (HR = 3.5 95% CI 1.2 to 9.9, p = 0.02) and the interaction between perceived stress and avoidance coping (HR = 7.0 per 5 unit increase for both scales, 95% CI 2.3 to 21.8, p = 0.003) were predictors of earlier relapse. CONCLUSIONS: In quiescent CD, a higher CRP, fistulising disease behaviour and disease confined to the colon were independent predictors of relapse. Moreover, patients under conditions of low stress and who scored low on avoidance coping (ie, did not engage in social diversion or distraction) were least likely to relapse. This study supports a biopsychosocial model of CD exacerbation.

Low urine pH is associated with reduced indinavir crystalluria in indinavir-treated HIV-infected individuals.

Indinavir is a potent HIV-1 protease inhibitor included in current antiretroviral therapeutic regimens. It is associated with renal and urological complications ascribed to indinavir crystalluria. We have previously reported that indinavir crystalluria is frequently observed soon after initiation of therapy. In a cohort of 54 asymptomatic indinavir-naive HIV-1-infected individuals during their first year of treatment with indinavir, approximately 25% of urinalyses (U/A) contained indinavir crystals. Because the determinants of the crystalluria are unknown, we examined the relationship between urine specific gravity (SG) and pH, singly and in combination, and indinavir crystalluria in these subjects. A total of 579 U/A were obtained from the study subjects at their scheduled monthly outpatient medical assessments. The frequency of indinavir crystalluria was lower in U/A with lower pH, irrespective of the SG. Conversely, U/A with high pH (> or = 6.0) had a higher frequency of indinavir crystalluria, which was further influenced by the urine SG. As a result, nearly half of the U/A (46.7%) with high pH (> or = 6.0) and intermediate-high SG (> or = 1.015) contained indinavir crystals. In conclusion, the frequency of indinavir crystalluria in asymptomatic HIV-1 infected individuals during their first year of treatment with indinavir was markedly influenced by the urine pH and SG. Our findings suggest that low urine pH may have a protective effect against indinavir crystalluria across the entire range of urine SG.

Early lymphocyte recovery following autologous peripheral stem cell transplantation is associated with better survival in younger patients with lymphoproliferative disorders.

Early absolute lymphocyte count (ALC) has become an important end point for engraftment in patients undergoing autologous peripheral stem cell transplantation (APSCT). In this retrospective study, we evaluate the prognostic significance of early recovery of ALC ( > or = 0.5 cells x 10(9)/l on or before day 15) following APSCT in predicting transplant outcome in 72 patients with lymphoproliferative disorders, including non-Hodgkin's lymphoma (n = 30), Hodgkin's lymphoma (n = 8) and multiple myeloma (n = 34). The median quantities of CD34+ stem cells and lymphocytes infused were 4.97 x 10(6)/kg (range 0.64-11.7) and 11.3 x 10(7)/kg (range 1.11-110) respectively. After a median follow-up of 18 months (range 2-68), 28 patients had experienced a relapse and 16 had died. Of the 72 patients, 27 (37%) demonstrated early recovery of ALC. Early recovery of ALC was strongly associated with long-term overall and disease-free survival in patients aged less than 50 years (P < 0.001). In both univariate and multivariate survival analyses, a shorter time from diagnosis to APSCT was associated with early recovery of ALC (P = 0.03). These findings indicate that early recovery of ALC may contribute to longer survival in younger patients with lymphoproliferative disorders. A shorter time from diagnosis to APSCT may favor recovery of ALC independent of the infused stem cell or lymphocyte doses.

Influence of RANTES, SDF-1 and TGF-beta levels on the value of interleukin-7 as a predictor of virological response in HIV-1-infected patients receiving double boosted protease inhibitor-based therapy.

OBJECTIVES: Interleukin-7 (IL-7), RANTES (regulated on activation, normal T cell expressed and secreted), stromal cell-derived factor-1 (SDF-1) and transforming growth factor-beta (TGF-beta) appear to share certain biological properties in vitro and all are involved in HIV-1 disease progression. Our earlier observations indicated that IL-7 levels decrease upon CD4 T-cell recovery and represent a new, independent predictor of virological response. Here, we examine associations among circulating levels of IL-7, RANTES, SDF-1 and TGF-beta in hopes of gaining insight into their contribution to the predictive value of IL-7. METHODS: Levels of IL-7, RANTES, SDF-1 and TGF-beta, and immune and viral parameters were assessed in HIV-1-infected patients. RESULTS: Cross-sectional (n=148) and longitudinal (n=36) analyses showed that levels of IL-7, but not RANTES, SDF-1 or TGF-beta, were increased in HIV-1-infected adults compared with those of healthy controls. In the cross-sectional study, levels of IL-7 were correlated with RANTES (r=0.31, P=0.002) and TGF-beta (r=0.53, P<0.001) but not with SDF-1 (r=0.12, P=0.22), and these associations were more pronounced in patients with CD4 T-cell counts >200 cells/microL. In contrast to IL-7, levels of RANTES, SDF-1 and TGF-beta were not correlated with CD4 T-cell counts. Longitudinal analysis revealed a marked decline in IL-7 levels accompanied by an increase in CD4 T-cell count following antiretroviral therapy (ART), but no changes in RANTES, SDF-1 or TGF-beta levels. Multivariate regression analysis showed no influence of baseline RANTES, SDF-1 or TGF-beta levels on the value of IL-7 as a predictor of virological response at 48 weeks. CONCLUSIONS: Collectively, these results indicate that changes in IL-7 levels did not induce changes in RANTES, SDF-1 or TGF-beta. Furthermore, they indicate that RANTES, SDF-1 or TGF-beta levels do not explain the predictor value of IL-7 in patients receiving ART.

Sample size requirements for case-control study designs.

BACKGROUND: Published formulas for case-control designs provide sample sizes required to determine that a given disease-exposure odds ratio is significantly different from one, adjusting for a potential confounder and possible interaction. RESULTS: The formulas are extended from one control per case to F controls per case and adjusted for a potential multi-category confounder in unmatched or matched designs. Interactive FORTRAN programs are described which compute the formulas. The effect of potential disease-exposure-confounder interaction may be explored. CONCLUSIONS: Software is now available for computing adjusted sample sizes for case-control designs.

Meta-analysis comparing newer antipsychotic drugs for the treatment of schizophrenia: evaluating the indirect approach.

BACKGROUND: Meta-analysis is a useful method to assess the efficacy of newer antipsychotic drugs compared with older drugs or placebo. However, few trials directly compare novel drugs to each other. OBJECTIVE: The purpose of this study was to evaluate the method of indirect meta-analysis by applying it to data on olanzapine versus haloperidol and risperidone versus haloperidol to enable a comparison between olanzapine and risperidone. METHODS: Published randomized controlled trials (RCTs) of risperidone, olanzapine, and/or haloperidol were identified through literature searches (1983 to 1999) of the MEDLINE, Current Contents, and HealthSTAR databases and reviewed. Data for the Brief Psychiatric Rating Scale (BPRS) total score, the Positive and Negative Syndrome Scale (PANSS) negative subscale, the percentage of patients using anticholinergic drugs, and the percentage of patients dropping out due to lack of efficacy, side effects, or any cause were extracted and combined using the indirect method. These findings were compared with those from a direct comparative study of olanzapine and risperidone. RESULTS: The literature search yielded 8 RCTs comparing risperidone to haloperidol and 3 comparing olanzapine to haloperidol. Only 1 trial directly comparing olanzapine and risperidone was found. In this trial, the change in BPRS total and PANSS negative subscale scores tended to be higher with olanzapine by 1.80 and 1.10, respectively, but these differences were not statistically significant. Indirect meta-analysis yielded similar results. Changes in both BPRS total scores and PANSS negative subscale scores tended to be higher with olanzapine by 0.37 and 0.54, respectively, and again, the differences were not statistically significant. In the indirect meta-analysis, the rate of anticholinergic drug use was 19.5% greater among patients treated with risperidone than among patients treated with olanzapine (P < 0.05). In the direct comparative RCT, the rate was 13.1% higher among patients treated with risperidone (P < 0.05). The dropout rates were similar for patients treated with risperidone and those treated with olanzapine in both analyses. CONCLUSION: An indirect meta-analysis of studies comparing olanzapine with haloperidol and risperidone with haloperidol yielded conclusions similar to those found in a direct comparative RCT of olanzapine and risperidone.

Prognostic markers in resectable non-small cell lung cancer: a multivariate analysis.

OBJECTIVE: To identify the prognostic significance of certain clinical, cellular and immunologic markers in resectable non-small cell lung cancer (NSCLC). DESIGN: A cohort of patients with resectable NSCLC was prospectively followed up for 8 years (100% follow-up). SETTING: A university hospital in a large Canadian city. PATIENTS: One hundred and thirteen consecutive patients who underwent surgical resection of primary NSCLC. MAIN OUTCOME MEASURES: Presence of peritumoral B lymphocytes (identified with antibody to CD20) and T lymphocytes (antibody to CD43), along with tumour markers (carcinoembryonic antigen [CEA], keratin, cytokeratin, S-100 protein, vimentin, chromogranin) and other factors such as age, sex, cell type, American Joint Committee on Cancer (AJCC) stage, histologic grade, DNA ploidy and S-phase fraction were correlated with survival. RESULTS: The mean age of patients in the study was 66.0 years; 60% were male. Histologic types of the tumours were: adenocarcinoma 57 (50.4%), squamous cell 47 (41.6%), adenosquamous 6 (5.3%) and large cell 3 (2.6%). AJCC stages were: I 66 (58.4%), II 20 (17.7%) and III 27 (23.9%). Histologic grades were: I (well differentiated) 31 (27.4%), II 50 (44.2%), III 29 (25.7%) and IV 3 (2.6%). Survival was 85% at 1 year (95% confidence interval [CI] 76%-90%), 44% at 5 years (95% CI 34%-53%) and 34% at 10 years (95% CI 22%-46%). Multivariate analyses using the Cox proportional hazards model for survival confirmed AJCC stage (p < 0.001) in all histologic subtypes to be the strongest factor of independent prognostic significance. It also revealed the presence of CD20-stained B lymphocytes (p = 0.04) in the peritumoral region of all tumours to be a positive prognostic factor. This relation was especially strong for nonsquamous cell carcinomas (p < 0.001). For squamous cell carcinomas, the immunohistochemical presence of CEA was of marginally negative prognostic value (p = 0.04). DNA ploidy and a high S-phase fraction showed no evidence of prognostic value for stage I tumours, but for stages II and III tumours there was strong evidence of prognostic value (p < 0.001 jointly). The evidence for DNA ploidy was especially strong in stages II and III squamous cell tumours (p = 0.008), and for a high S-phase fraction was strongest in stages II and III nonsquamous cell tumours (p = 0.002). CONCLUSIONS: AJCC stage remains the most important prognostic indicator from a variety of clinical variables and tumour markers in postoperative patients with resectable NSCLC. For nonsquamous cell lung carcinomas, the presence of peritumoral B lymphocytes was strongly associated with improved survival, suggesting an important role for humoral mediated immunity.

Clinical, biological, and histologic parameters as predictors of relapse in ulcerative colitis.

BACKGROUND & AIMS: Ulcerative colitis (UC) is a chronic relapsing inflammatory bowel disease. We aimed to assess whether clinical, biological, and histologic parameters in quiescent UC predict time to clinical relapse. METHODS: Seventy-four patients with clinically and endoscopically determined inactive UC were followed up for 1 year or for a shorter period if they had a relapse. Serum erythrocyte sedimentation rate; C-reactive protein, interleukin (IL)-1beta, IL-6, and IL-15 values; anti-neutrophil cytoplasmic antibody titers; and rectal biopsy specimens were obtained at baseline, at 6 and 12 months, and/or at relapse. Multivariate survival analysis was performed to determine independent predictors of clinical relapse. RESULTS: Twenty-seven patients relapsed (19/42 women; 8/32 men). Multivariate Cox regression analysis retained younger age (P = 0.003; hazard ratio, 0.4 per decade), greater number of prior relapses in women (P < 0.001; hazard ratio, 1.6 per prior relapse), and basal plasmacytosis (P = 0.003; hazard ratio, 4.5) on rectal biopsy specimens as predictors of shorter time to clinical relapse. Kaplan-Meier survival curves showed the 20-30-year-old age group and women with more than 5 prior relapses to be groups with shorter times to relapse. CONCLUSIONS: Younger age, multiple previous relapses (for women), and basal plasmacytosis on rectal biopsy specimens were independent predictors of earlier relapse. These findings may help identify patients with inactive UC who will require optimal maintenance medical therapy.

The generalization of the odds ratio, risk ratio and risk difference to r x k tables.

Familiar measures of association for 2 x 2 tables are the odds ratio, the risk ratio and the risk difference. Analagous measures of outcome-exposure association are desirable when there are several degrees of severity of both exposure and disease outcome. One such measure (alpha), which we label the general odds ratio (OR(G)), was proposed by Agresti. Convenient methods are given for calculation of both standard error and 95 per cent confidence intervals for OR(G). Other approaches to generalizing the odds ratio entail fitting statistical models which might not fit the data, and cannot handle some zero frequencies. We propose a generalization of the risk ratio (RR(G)) following the statistical approaches of Agresti, Goodman and Kruskal. A method of calculating the standard error and 95 per cent confidence interval for RR(G) is provided. A known statistic, Somers' d, fulfils the characteristics necessary for a generalized risk difference (RD(G)). These measures have straightforward interpretations, are easily computed, are at least as precise as other methods and do not require fitting statistical models to the data. We also examine the pooling of such measures as in, for example, meta-analysis.

GEE analysis of negatively correlated binary responses: a caution.

The method of generalized estimating equations has become almost standard for analysing longitudinal and other correlated response data. However, we have found that if binary responses have less than binomial variation over clusters, and are modelled using exchangeable correlations, prevailing software implementations may give unreliable results. Bounding the negative correlation away from its theoretical minimum may not always be a satisfactory solution. In such instances, using the independence working correlation structure and robust SEs is a more trustworthy alternative.

A physician-centred intervention to shorten hospital stay: a pilot study.

BACKGROUND: Studies of length of stay (LOS) in hospital usually focus on physician-independent factors. In this study, the authors identified physician-dependent factors and tested an intervention aimed at them to determine its effect on LOS. METHODS: A prospective comparison of LOS on 2 general medical wards in a tertiary care teaching hospital before and after the intervention. The pre-intervention (control) period and the intervention period were each 4 weeks. The intervention consisted of a checklist for planning management and discharge. RESULTS: Overall, the mean LOS was shorter during the intervention period than during the control period, but the difference was not statistically significant (12.0 and 14.4 days respectively, p = 0.13). The difference was significant on ward A (11.0 v. 14.7 days respectively, p = 0.02) but not on ward B (13.0 and 14.0 days respectively, p = 0.90). INTERPRETATION: An intervention at the level of the admitting physician may help to shorten LOS on a general medical ward.

External comparisons from nested case-control designs.

The nested case-control design, used to sample within cohorts, is usually employed for internal comparisons. We propose to use this design for external comparisons. We present two probability-weighted estimators of the expected number of cases under a given exposure, based on external rates, for two versions of the nested case-control design. These estimators are used, along with their variance estimators, to form confidence intervals for standardized mortality ratios. The estimators are practically unbiased, whereas the naive estimator that treats the nested case-control sample as a random sample of the cohort is clearly biased. An estimator from the alternative Cox model-based approach is found to be substantially biased when applied in this context. Comparing the proposed estimators for nested case-control designs to a corresponding estimator for the case-cohort design, we found that the correlation between follow-up time and exposure time (that is, the amount of time under the exposure effect) has an impact on which type of design is more efficient for external comparisons. A small correlation favors the case-cohort design and a large correlation the nested case-control design. We examine empirical properties of these estimators through computer simulations, using a cohort study of the incidence of second cancer in 2,189 patients with Hodgkin's disease.

Adjusted odds ratios for case-control studies with missing confounder data in controls.

Nonexperimental studies using computerized databases often give rise to missing or partially available information on confounders. A frequent situation occurs when data on exposure are available for all subjects of a case-control study, but data on confounders are available only for the cases but not for the controls. In that situation, the fact of confounding can be verified by assessing the association between exposure and a confounder in the cases, but the data are insufficient to produce an adjusted estimate of the relative risk if confounding is found to be present. We propose simple conditions under which an adjusted estimate of the relative risk can be obtained when data on a confounder are available only for the cases, and we derive formulae for the estimator and its confidence limits. The method requires an external estimate of the confounder prevalence or, additionally, of the confounder-exposure odds ratio. We illustrate the technique with data from a nested case-control study of the risk of acute cardiac death associated with the use of bronchodilator drugs within a cohort of 12,301 asthmatics, with smoking as the confounder of interest.

A confidence interval for Pr(X < Y) - Pr(X > Y) estimated from simple cluster samples.

Distribution-free confidence intervals based on Somers' d (a simple function of the Mann-Whitney U statistic) are developed for Pr(X < Y) - Pr(X > Y), where X and Y are the ranks of any two observations from two independent populations. The approach accommodates a complex sampling design, and explicit formulas are given for simple cluster sampling. The method also accommodates simple progressive left- and right-censoring. The accuracy of the interval is shown to improve when the tanh-1 transform is used. A bootstrap solution was tried and did not perform as well as the proposed solution.