EUCLID: an outcome analysis tool for high-dimensional clinical studies

Treatment management decisions in three-dimensional conformal radiation therapy (3DCRT) and intensity-modulated radiation therapy (IMRT) are usually made based on the dose distributions in the target and surrounding normal tissue. These decisions may include, for example, the choice of one treatment over another and the level of tumour dose escalation. Furthermore, biological predictors such as tumour control probability (TCP) and normal tissue complication probability (NTCP), whose parameters available in the literature are only population-based estimates, are often used to assess and compare plans. However, a number of other clinical, biological and physiological factors also affect the outcome of radiotherapy treatment and are often not considered in the treatment planning and evaluation process. A statistical outcome analysis tool, EUCLID, for direct use by radiation oncologists and medical physicists was developed. The tool builds a mathematical model to predict an outcome probability based on a large number of clinical, biological, physiological and dosimetric factors. EUCLID can first analyse a large set of patients, such as from a clinical trial, to derive regression correlation coefficients between these factors and a given outcome. It can then apply such a model to an individual patient at the time of treatment to derive the probability of that outcome, allowing the physician to individualize the treatment based on medical evidence that encompasses a wide range of factors. The software's flexibility allows the clinicians to explore several avenues to select the best predictors of a given outcome. Its link to record-and-verify systems and data spreadsheets allows for a rapid and practical data collection and manipulation. A wide range of statistical information about the study population, including demographics and correlations between different factors, is available. A large number of one- and two-dimensional plots, histograms and survival curves allow for an easy visual analysis of the population. Several visual and analytical methods are available to quantify the predictive power of the multivariate regression model. The EUCLID tool can be readily integrated with treatment planning and record-and-verify systems.     

Individual predictors of outcome in group treatment for incest survivors

Sixty-five abused women who participated in time-limited group therapy were evaluated before treatment on demographic variables, sexual abuse history, characteristics of the family of origin, and initial levels of depression and distress. Hierarchical regression analyses were conducted to determine the impact of these variables on response to treatment. Results suggest that education, marital status, type of sexual contact, and initial levels of depression and distress predicted response to treatment. In addition, a history of previous therapy showed a tendency to interact with type of group format to predict outcome. Implications for therapy and further research are considered.     

Long-term predictors for psychological outcome of pre-symptomatic testing for late-onset neurological diseases

This longitudinal study aimed at determining predicting variables for middle and long-term psychological disturbance due pre-symptomatic testing (PST) for two late-onset neurological diseases, Huntington disease (HD) and TTR (transthyretin protein) familial amyloid polyneuropathy (FAP) Val30Met (now classified as Val50Met). 196 clinical records of persons who performed PST at least three years ago and answered to the two stages of evaluation (before PST and least 3 years after disclosure of results) were analysed. For this purpose, regression analysis was performed, showing that the Positive Symptom Distress Index (PSDI), psychoticism, somatization and paranoid ideation dimensions assume predictive value in the middle and long-term impact for total anxiety and PSDI. The result of PST was not a relevant predictor. The application of an evaluation instrument of various psychopathological dimensions played a fundamental role in the detection of clinical situations that may arise several years later after PST. Attention should be paid to providing psychological support to persons at-risk who, at the pre-test phase, present some psychopathology indices before pursuing with genetic testing.     

The predictors of outcome in total knee arthroplasty for post-traumatic arthritis

BackgroundThe outcomes of total knee arthroplasty (TKA) for post-traumatic arthritis are less reliable than for idiopathic osteoarthritis. These patients tend to be younger, present with varying degrees of deformity, and often have a history of prior surgery, resulting in a wide spectrum of pathology. We hypothesized that preoperative variables, in particular the location of deformity, would predict pain and functional outcomes.MethodsThe outcomes of total knee arthroplasty for post-traumatic arthrosis were studied in 47 knees treated at our institution. All patients were administered the Knee Society Score (KSS) preoperatively and at follow-up (mean 52 months, range 16 to 124). We classified the defects into four categories: intra-articular, metaphyseal, diaphyseal, and combined femoral and tibial deformities.ResultsThere was a significant improvement in KSS scores for pain (p < 0.0001) and a trend toward higher function (p = 0.06) comparing preoperative scores to final follow-up. The largest improvements in pain and functional scores were in patients with isolated articular deformities, while patients with combined tibial and femoral deformities did not have significant improvements in pain or function. Soft-tissue defects requiring flap coverage were associated with worsening in the pain score (p = 0.027).ConclusionsThe location of post-traumatic deformity and compromise of the soft-tissue envelope influence the pain and functional outcomes of total knee arthroplasty for post-traumatic arthritis. Specifically, isolated articular deformities have the largest improvement in pain and function while patients with combined tibial and femoral deformities as well as patients with soft-tissue compromise experience poor outcomes.Level of evidenceIV—Retrospective Case Series.     

Cardiac troponins: outcome predictors in hemodialysis patients

Cardiovascular diseases represent the main cause of death in hemodialysis (HD) patients. Cardiac troponins (cTnT and cTnI) are indicators of myocardial damage. The aims of this study were to assess the prevalence of increased serum cTn in the absence of acute coronary syndrome, to determine overall and cardiovascular mortality rates, and to investigate the possible predictive values of cTnT and cTnI on the outcome in HD patients over a 2-year follow-up period. The study included 115 patients (71 men and 44 women) with an average age of 53.30 ± 12.17 years who had undergone regular HD for 4.51 ± 4.01 years and had a mean HD adequacy (Kt/Vsp) of 1.17 ± 0.23. Increased serum cTnT concentration was found in 37.39% of patients and elevated serum cTnI concentration was present in 11.30% of HD patients without symptoms or signs of acute coronary syndrome. The average 2-year mortality rate was 13.74% and the average 2-year cardiovascular mortality rate was 8.51%. Patients with serum cTnT levels greater than 0.10 ng/ml had significantly lower overall and cardiovascular survival rates than patients with serum cTnT levels of less than 0.10 ng/ml. Patients with serum cTnI levels greater than 0.15 ng/ml had significantly lower overall and cardiovascular survival rates than patients with serum cTnI of less than 0.15 ng/ml. In patients on regular HD, cTn levels are significant outcome predictors.     

Depression and anxiety as predictors of outcome after myocardial infarction

OBJECTIVE: The objective of this study was to investigate the significance of emotional distress immediately after a myocardial infarction as a predictor of physical, psychological, and social outcomes and resource use. METHODS: In an epidemiological survey, demographic and cardiological data were obtained for all patients from a defined geographical area who had had a myocardial infarction (according to diagnostic criteria of the Monitoring Trends and Determinants in Cardiovascular Disease [MONICA] trial). Hospital survivors were interviewed and were asked to complete self-report assessments on mental state and quality of life. Full replies were available at baseline for 347 subjects. Self-report follow-up questionnaire information was collected 3 months and 1 year later. RESULTS: Fifteen percent of patients scored as probable cases of anxiety or depression. They were more likely than noncases to report preinfarct distress and poor adjustment (as indicated on the 36-item Medical Outcome Study short form). There was an improvement at 3 months, but there was little overall or individual change after that time. Anxiety and depression did not predict subsequent mortality but did significantly predict poor outcome at 1 year on all dimensions of the 36-item short form quality-of-life measure and on specific measures of everyday activity and reports of chest pain, use of primary care resources, and secondary prevention lifestyle changes. CONCLUSIONS: Subjects who are distressed in the hospital are at high risk of adverse psychological and quality-of-life outcomes during the ensuing year. Our findings strengthen the argument for in-hospital identification and treatment of patients with depression and anxiety after myocardial infarction.     

Clinical predictors and outcome of bowel resection in paediatric intussusception

IntroductionSurgery remains the mainstay in treating intussusception in developing countries, with a correspondingly high bowel resection rate despite a shift to non-operative reduction in high-income countries.ObjectiveTo assess factors associated with bowel resection and the outcomes of resection in childhood intussusception.MethodsA review of children with intussusception between January 2006 and December 2015 at the University College Hospital, Ibadan, Nigeria. The patients were categorized based on the need for bowel resection and analysis done using the SPSS version 23.Results121 children were managed for intussusception during this period. 53 (43.8%) had bowel resection, 61 (50.4%) did not require resection and 7 (5.8%) were unknown. 40 (75.5%) of the resections were right hemi-colectomy. The presence of fever, abdominal pain, distension, rectal mass, age < 12 months, heart rate > 145/min and duration of symptoms > 2 days were associated with the need for bowel resection (p < 0.05). However, only age and abdominal pain independently predicted need for resection. Bowel resection was more associated with development of post-operative complications and prolonged hospital stay (p < 0.05).ConclusionInfants presenting with abdominal pain and abdominal distension after two days of onset of symptoms were more likely to require bowel resection. Resection in intussusception significantly increased post-operative complications and length of hospital stay.     

Electrodermal predictors of functional outcome and negative symptoms in schizophrenia

The ability of electrodermal variables to predict negative symptoms and functional outcome over a 1-year period in schizophrenia was investigated in 78 young, recent-onset outpatients. Patients were stabilized on standardized medication and largely free of psychotic symptoms. Higher levels of both tonic (skin conductance level, nonspecific skin conductance response rate) and phasic (number of skin conductance orienting responses) activity were associated with more negative symptoms and with a combination of poorer social and occupational outcome at 1-year follow-up. This pattern was seen in both male and female patients, and in older and younger patients. Results are interpreted as suggesting that high levels of arousal and overreactivity to the environment may interfere with efficient cognitive processing in schizophrenia, contributing to poor outcome, and that negative symptoms might partially serve as a means of coping with overarousal.     

Glomerulonephritis with crescents among adult Saudi patients outcome and its predictors

The aim of this study is to investigate the clinical and pathological features and outcome of glomerulonephritis with crescents among adult patients. This is a retrospective study of all cases of crescentic GN seen over a 9-year period (2001–2010). Histological features were assessed, and renal function at baseline and end of follow-up period was recorded. Results among different etiological groups at baseline and end of follow-up period were compared. The mean age in the whole group was 35.6 years (16.2), with the lowest mean in the lupus nephritis (LN) group [27.7 years (9.9)] and the highest in the pauciimmune glomerulonephritis (PIGN) group (P = 0.001). There were 72 cases enrolled in the study. LN accounted for 49.3% of the cases, PIGN for 26.5%, other immune complex glomerulonephritis (ICGN) for 19% and post-infectious GN accounted for 6.3% The majority (85.7%) of the patients had renal impairment at presentation (mean serum creatinine levels were 247 (85) μmol/l, 412 (75) μmol/l and 230 (141) μmol/l in LN, PICN and ICGN, respectively (P = 0.05). Women accounted for 85.3, 76.5 and 36.2% of the patients in LN, PICN and ICGN, respectively (P = 0.025). By the end of the follow-up period of 26 (22.9) months, 25.8% of the patients were requiring dialysis (16.70% in the LN group, 50% in PIGN and 25% in ICGN (P = 0.05) and 21.7% had nephrotic range proteinuria (16.7, 1 and 33.3%, respectively (P = 0.4). Using logistic multivariate analysis, the only independent factors found to predict need for dialysis of prognosis were percent of sclerosed glomeruli (P = 0.05) and presence of ATN (P = 0.028). Baseline proteinuria or SCr, gender and number of glomeruli with crescents, on the other hand, did not impact prognosis. Using linear regression multivariate analysis, SCr, protein excretion and activity score at biopsy did not influence change in SCr or final SCr during the follow-up period. Using ANOVA to compare the groups of LN, PIGN and ICGN), we found significant differences only in gender between LN and ICGN (P = 0.035), in percent glomerular global sclerosis (between LN and PIGN (P = 0.007) and between LN and ICGN (P = 0.012) and in age (between LN and PIGN (P = 0.006). Almost half of our patients with CrGN were due to LN which is higher than that reported by others where PIGN was the more prevalent etiology. Patients with PICN were older and had worse prognosis. This could be explained by the higher number of globally sclerosed glomeruli in the PIGN group.     

Mapping high-dimensional data onto a relative distance plane--an exact method for visualizing and characterizing high-dimensional patterns

We introduce a distance (similarity)-based mapping for the visualization of high-dimensional patterns and their relative relationships. The mapping preserves exactly the original distances between points with respect to any two reference patterns in a special two-dimensional coordinate system, the relative distance plane (RDP). As only a single calculation of a distance matrix is required, this method is computationally efficient, an essential requirement for any exploratory data analysis. The data visualization afforded by this representation permits a rapid assessment of class pattern distributions. In particular, we can determine with a simple statistical test whether both training and validation sets of a 2-class, high-dimensional dataset derive from the same class distributions. We can explore any dataset in detail by identifying the subset of reference pairs whose members belong to different classes, cycling through this subset, and for each pair, mapping the remaining patterns. These multiple viewpoints facilitate the identification and confirmation of outliers. We demonstrate the effectiveness of this method on several complex biomedical datasets. Because of its efficiency, effectiveness, and versatility, one may use the RDP representation as an initial, data mining exploration that precedes classification by some classifier. Once final enhancements to the RDP mapping software are completed, we plan to make it freely available to researchers.     

Patient expectations and therapeutic alliance as predictors of outcome in group cognitive-behavioral therapy for insomnia

Despite growing evidence for the efficacy of cognitive-behavioral therapy for insomnia (CBT-I), few data exist on the relation between process and outcome for this treatment. Drawing on interpersonal theory and the broader psychotherapy literature, this study examined the contribution of patient expectations and the therapeutic alliance to outcomes in group CBT-I. For patients with low early treatment expectations for improvement, those perceiving the therapist as higher in affiliation had greater reduction in sleep problems. Perceiving the therapist as critically confrontive was generally associated with less treatment satisfaction, and particularly so for those individuals who came to treatment with high expectations for improvement. Critical confrontation also differentiated dropouts from continuers, with dropouts experiencing their therapist as more critically confrontive.     

Pretreatment and process predictors of outcome in interpersonal and cognitive behavioral psychotherapy for binge eating disorder

The present study examined pretreatment and process predictors of individual nonresponse to psychological group treatment of binge eating disorder (BED). In a randomized trial, 162 overweight patients with BED were treated with either group cognitive-behavioral therapy or group interpersonal psychotherapy. Treatment nonresponse, which was defined as nonabstinence from binge eating, was assessed at posttreatment and at 1 year following treatment completion. Using 4 signal detection analyses, greater extent of interpersonal problems prior to treatment or at midtreatment were identified as predictors of nonresponse, both at posttreatment and at 1-year follow-up. Greater pretreatment and midtreatment concerns about shape and weight, among those patients with low interpersonal problems, were predictive of posttreatment nonresponse. Lower group cohesion during the early treatment phase predicted nonresponse at 1-year follow-up. Attention to specific pre- or intreatment predictors could allow for targeted selection into differential or augmented care and could thus improve response to group psychotherapy for BED.     

Cognitive therapy of depression: pretreatment patient predictors of outcome

This review examines the role of patient predictors of outcome in cognitive therapy of depression. Studies that meet eligibility criteria are reviewed for demonstrated linkage between various predictors (i.e., pretreatment severity, historical features, demographic predictors, dysfunctional attitudes and other cognitive features, and treatment acceptability) and outcome, and several effects are found. Notably, high pretreatment severity scores are associated with poorer response to cognitive therapy, as are high chronicity, younger age at onset, an increased number of previous episodes, and marital status. High pretreatment levels of dysfunctional attitudes and certain beliefs about the nature of depression were also found to predict differential response to cognitive therapy of depression. Limitations of the research and directions for further investigations of patient predictors of outcome in cognitive therapy of depression are provided.     

Cytokine production and monocyte HLA-DR expression as predictors of outcome for patients with community-acquired severe infections

This study was performed to evaluate the impact of pro- and anti-inflammatory molecules and human leukocyte antigen DR (HLA-DR) expression as markers of immune status for the final outcome of septic patients. The study included 30 patients with severe sepsis due to community-acquired infections. Concentrations of tumor necrosis factor alpha (TNF-alpha), interleukin-6 (IL-6), IL-8, IL-10, and transforming growth factor beta1 (TGF-beta1) in serum, as well as monocyte HLA-DR expression, were determined on admission and on days 3, 10, 13, and 17 during hospitalization. Of the 30 patients enrolled, 13 survived, while 17 died during their hospital stay. All patients had significantly lower HLA-DR expression and higher pro- and anti-inflammatory cytokine levels than healthy individuals. HLA-DR expression was significantly decreased in nonsurvivors at almost all time points. In nonsurvivors, higher levels in serum of TNF-alpha on days 13 and 17; IL-6 levels on day 3; and IL-10 on days 3, 10, and 13 were found. Baseline levels of TGF-beta1 were significantly higher in survivors. Independent risk factors of mortality were IL-10 levels on days 3 and 10, while monocyte HLA-DR expression on admission was a good predictor for survival. Several pro- and anti-inflammatory cytokines are oversynthesized during severe infections, especially in patients with a poor outcome. Monocyte HLA-DR expression is an early and constant predictive marker for survival in severe sepsis, while serum IL-10 levels on days 3 and 10 have negative prognostic value for the final outcome.     

Rorschach predictors of therapeutic outcome for inpatient treatment of children: a proactive study

The present study tested the hypothesis that Rorschach measures of psychological instability and perceptual sensitivity (i.e., es, es-EA, Lambda, Blends, Zf, & Zsum) would be sensitive to therapeutic change after inpatient treatment of children. Twelve male children were administered the Rorschach Ink Blot test in accordance with Exner's (1986) procedures and comprehensive system, pretreatment and posttreatment. The 12 children received inpatient treatment an average of 12 months, and all children showed improved functioning after treatment. Pre- to posttreatment Rorschach performance changes provided a partial confirmation of the hypothesis. Specifically, significant changes in es and Lambda were found posttreatment. As such, these Rorschach variables appear to be sensitive to therapeutic change in children.