Kaplan-Meier Survival Distributions Research Articles

MRI-based radiomics approach for the prediction of recurrence-free survival in triple-negative breast cancer after breast-conserving surgery or mastectomy.

To explore the value of a radiomics signature and develop a nomogram combined with a radiomics signature and clinical factors for predicting recurrence-free survival in triple-negative breast cancer patients. We enrolled 151 patients from the cancer imaging archive who underwent preoperative contrast-enhanced magnetic resonance imaging. They were assigned to training, validation and external validation cohorts. Image features with coefficients not equal to zero in the 10-fold cross-validation were selected to generate a radiomics signature. Based on the optimal cutoff value of the radiomics signature determined by maximally selected log-rank statistics, patients were stratified into high- and low-risk groups in the training and validation cohorts. Kaplan-Meier survival analysis was performed for both groups. Kaplan-Meier survival distributions in these groups were compared using log-rank tests. Univariate and multivariate Cox regression analyses were used to construct clinical and combined models. Concordance index was used to assess the predictive performance of the 3 models. Calibration of the combined model was assessed using calibration curves. Four image features were selected to generate the radiomics signature. The Kaplan-Meier survival distributions of patients in the 2 groups were significantly different in the training (P < .001) and validation cohorts (P = .001). The C-indices of the radiomics model, clinical model, and combined model in the training and validation cohorts were 0.772, 0.700, 0.878, and 0.744, 0.574, 0.777, respectively. The C-indices of the radiomics model, clinical model, and combined model in the external validation cohort were 0.778, 0.733, 0.822, respectively. The calibration curves of the combined model showed good calibration. The radiomics signature can predict recurrence-free survival of patients with triple-negative breast cancer and improve the predictive performance of the clinical model.

551 Killip class predictors and prognostic role in acute myocardial infarction

Abstract Aims Killip classification is a simple and fast clinical tool for risk stratification in patients with acute coronary syndrome (ACS). However, predictors of high Killip class at admission and its prognostic impact in the clinical contest of myocardial infarction with nonobstructive coronary artery (MINOCA) are still poorly known. To identify the clinical predictors of high Killip class and its potential prognostic role on in-hospital and follow-up outcomes in patients with MINOCA compared to patients with myocardial infarction with obstructive coronary artery (MIOCA). Methods and results We included all consecutive patients with myocardial infarction (MI) undergoing coronary angiogram between 2016 and 2019 at our hospital. According to 2016 ESC Position Paper criteria, we considered as MINOCA all patients with acute MI and with the angiographic conventional cut-off of &amp;lt; 50% coronary stenosis without clinically apparent alternative diagnosis (e.g. sepsis, stroke, pulmonary embolism, myocarditis, and Tako-tsubo). We analysed Killip class of MINOCA patients comparing with those of MIOCA (coronary stenosis ≥50%). Kaplan–Meier (KM) curves were developed for the comparison of overall-mortality among MINOCA with high Killip class (major than 1) compared to other. Multivariate logistic regression analysis was used to determine the predictors of high Killip class both in the MINOCA and MIOCA populations. Among 3165 MI, 260 patients fulfilled the 2016 ESC criteria for MINOCA. Overall, 62.3% were males and the mean age was 68.6 ± 13.2 years. The median follow-up time was 23.3 ± 14.5 months. Killip class &amp;gt;1 occurred in 24 patients in MINOCA group and 507 in MIOCA group (17.5% vs. 9.2%, P = 0.001). The KM survival distributions were significantly different across Killip class &amp;gt;1 (P &amp;lt; 0.001) in both populations with higher mortality in patients with higher Killip class. Finally, the multivariate logistic regression showed that the predictors of high Killip class at time of presentation in MIOCA population were older age [odds ratio: 1.04, 95% CI: (1.03–1.06), P &amp;lt; 0.001], diabetes [odd ratio 0.63, 95% CI (0.48–0.81), P &amp;lt; 0.001], ST elevation [odds ratio: 0.65, 95% CI (0.48–0.89), P = 0.008], left ventricle ejection fraction [odds ratio: 0.95, 95% CI (0.94–0.96), P &amp;lt; 0.001], and elevated cardiac troponin [odds ratio: 1.00, 95% CI (1.00–1.00), P = 0.01]. Older age [odds ratio: 1.08, 95% CI (1.03–1.14), P = 0.003], ST elevation [odd ratio 0.14, 95% CI (0.02–0.93), P = 0.042], and diabetes [odd ratio 3.60, 95% CI (1.08–1.96), P = 0.037] were predictors of high Killip class in MINOCA, however left ventricle ejection fraction (P = 0.3) and elevated cardiac troponin (P = 0.6) did not predict the high Killip class in MINOCA patients. Conclusions Our data suggest that Killip classification performed at the time of admission is a useful clinical marker of a high risk of early and late adverse cardiovascular events even in patients with MINOCA. The predictors of the high Killip class at time of presentation in MIOCA were older age, diabetes, ST elevation, left ventricle ejection fraction, and elevated cardiac troponin. Older age, ST elevation, and diabetes were predictors of high Killip class even in MINOCA, however left ventricle ejection fraction and elevated cardiac troponin did not predict the high Killip class in MINOCA patients. These results could reflect the different pathogenetic myocardial damage in MINOCA and MIOCA populations. Further studies are needed to evaluate these pathological mechanisms.

Cardiac MRI risk factors predict atrial arrhythmia development in repaired tetralogy of Fallot patients

Abstract Background/Introduction Tetralogy of Fallot (TOF) patients with atrial arrhythmias (AA) experience greater morbidity and mortality than do their arrhythmia-free counterparts. Cardiac MRI (cMRI) has proven useful in elucidating risk factors for arrhythmia development in this population, but more work needs to be done to assess the efficacy of predicting AA onset. Purpose To evaluate demographic and clinical metrics, including those routinely assessed on cMRI, that predict AA onset in repaired TOF. Methods 175 repaired TOF patients ≥18 years of age and without prior AA undergoing initial cMRI between 2003 and 2020 were identified from a tertiary U.S. hospital database. Subsequent development of AA (specifically atrial fibrillation, atrial flutter, and atrial tachycardia) was evaluated by chart review and device checks (if present). Kaplan-Meier analysis was used to assess arrhythmia-free survival based on cMRI characteristics, and log-rank test used to compare differences in survival curves. Cutoff values for Kaplan-Meier analysis were the mean value (for the respective variable being measured) of our cohort at the time of their initial cMRI. Results Over a median follow-up of 3.3 years, 30 patients (17%) developed AA, including 16 with atrial fibrillation (53%), 20 with atrial flutter (67%), and 2 with atrial tachycardia (7%). Age at cMRI (p&amp;lt;0.01), history of hypertension (p=0.01), and time from birth to complete TOF repair (p&amp;lt;0.05) predicted AA. cMRI predictors included indexed right ventricular (RV) end diastolic and systolic volumes (both p&amp;lt;0.01), as well as indexed right atrial (RA) volume (p&amp;lt;0.05) and pulmonic regurgitation (PR) fraction (p&amp;lt;0.05). Indexed RV and RA volumes but not left sided chamber volumes predicted the development of atrial fibrillation. All Kaplan-Meier survival distributions were significantly different with the exception of those for right atrial volumes. Conclusions Our study suggests that RA and RV volumes may play a valuable role in predicting atrial arrhythmia onset in the repaired TOF population. Disease progression in a predominantly right-sided heart disease may play a role in predicting the onset of both “right sided” arrhythmias such as atrial flutter, as well as “left sided” arrhythmias such as atrial fibrillation. Additionally, the increased AA burden in patients with mild RV dilation seen in our study may suggest a potential benefit to early pulmonic valve replacement in this patient group. Funding Acknowledgement Type of funding sources: None. Clinical and cMRI CharacteristicsKaplan-Meier Survival Analysis

Characteristics and Outcomes in Pediatric Non-Central Nervous System Malignant Rhabdoid Tumors: A Report from the National Cancer Database

Pediatric non-central nervous system (CNS) malignant rhabdoid tumors (MRTs) are rare and aggressive malignancies without standard treatment strategies. The National Cancer Database (NCDB) was utilized to describe the incidence, characteristics, treatment strategies, and outcomes in pediatric patients. Patients <18 years of age and diagnosed with non-CNS MRTs were analyzed from the NCDB from 2004 to 2014. Log-rank tests compared differences in Kaplan-Meier survival distributions. Univariate and multivariable Cox proportional hazard regression models identified predictors of mortality. Overall, 202 patients were identified. Soft tissue tumors were most common and metastatic disease was present at diagnosis in 34.2% of patients. The 1- and 5-year overall survival (OS) rates were 48.8% and 35.9%, respectively. Multivariable analysis revealed that age <1 year and the presence of metastasis were negative prognostic indicators (p=0.058). The 1- and 5-year OS rates were 59.9% and 46.5%, respectively, for patients who received surgical intervention (n=143) compared with 12.3% and 7.4%, respectively, for those treated nonoperatively (n=59; p<0.01). Surgical resection was associated with improved outcomes on univariate analysis, although it was no longer an independent predictor of survival on multivariate analysis (p=0.18). In the cohort of surgical patients, the presence of residual disease trended towards clinically significant worse outcomes (p=0.13). Patients with non-CNS MRTs who were diagnosed in infancy and had metastatic disease had worse survival outcomes. Although surgical resection was associated with improved survival in non-CNS MRTs, it was not independently associated with survival on multivariate analysis. Efforts to improve survival may instead depend on improving chemotherapeutic strategies and developing targeted therapies.

On-treatment immune prognostic score for patients with relapsed and/or metastatic head and neck squamous cell carcinoma treated with immunotherapy

BackgroundPrevious studies have suggested that inflammatory markers (neutrophil-to-lymphocyte ratio (NLR), lactate dehydrogenase (LDH) and fibrinogen) are prognostic biomarkers in patients with a variety of solid cancers, including those treated with...

The Clinical Impact of Psoas Muscle Cross-Sectional Area on Medium-Term Mortality After Transcatheter Aortic Valve Implantation

Psoas muscle cross-sectional area (CSA) is a proposed marker of frailty associated with mortality after transcatheter aortic valve implantation (TAVI). We assessed the impact of psoas CSA on medium-term mortality over 5 years in a large cohort, adjusted for pre-procedural variables. This single-centre registry-derived analysis assessed 1,731 consecutive TAVI patients between 2007 and 31 April 2015 with available abdominal computed tomography scans. Sex-stratified, height-adjusted psoas CSA was measured mid-body of the fourth lumbar vertebra. Kaplan-Meier survival distributions across psoas CSA quartiles were compared. Cox and logistic regression models were used to assess baseline variables associated with the primary outcome, which was mortality within 5 years. Median age was 81 years (interquartile range, 77 - 85); 52.5% were women. The primary endpoint occurred in 555 patients over a mean follow-up of 775 days. Lower psoas CSA quartile patients were older, had a lower body mass index, lower creatinine clearance, and lower rates of previous cardiac surgery, with higher rates of diabetes, coronary artery disease, pacemaker, anaemia, hypoalbuminaemia, and higher European System for Cardiac Operative Risk Evaluation (EuroSCORE). Unadjusted survival by psoas CSA quartile was significantly different in men (log rank p=0.041) but not women (p=0.099). In Bonferroni-adjusted multivariate analysis, psoas CSA quartiles were not significantly associated with mortality. Hypoalbuminaemia (hazard ratio [HR], 2.10; 95% confidence interval [CI], 1.53 - 2.87 [p<0.001]) and increasing age (HR, 1.03 per year; 95% CI, 1.01 - 1.05 [p=0.002]) were associated with increased risk; female sex (HR, 0.63; 95% CI 0.51 - 0.78 [p<0.001]), and hypercholesterolaemia (HR, 0.67; 95%, CI 0.54 0.83 [p<0.001]) with reduced risk. Psoas CSA was not significantly associated with mortality after adjusting for pre-procedural variables. Hypoalbuminaemia, sex, hypercholesterolaemia, and age were significantly associated with mortality after TAVI.

Oral fingolimod for chronic inflammatory demyelinating polyradiculoneuropathy (FORCIDP Trial): a double-blind, multicentre, randomised controlled trial

Fingolimod is approved for the treatment of relapsing-remitting multiple sclerosis and was effective in experimental autoimmune neuritis in rats, a possible model for chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). We aimed to evaluate the efficacy of fingolimod in delaying disability progression in patients with CIDP who withdrew from currently effective treatments (intravenous immunoglobulin [IVIg] or corticosteroids). This double-blind, multicentre, randomised, placebo-controlled, parallel-group, event-driven study was done at 48 neurology centres in Australia, Canada, Israel, Japan, the USA, and nine countries in Europe. Participants with CIDP who were receiving IVIg or corticosteroids were randomly assigned (1:1) to once-daily oral fingolimod 0·5 mg or placebo. Owing to the event-driven design, treatment duration was flexible and could be up to 4·5 years. Randomisation was done with an automated interactive voice response-web response system and was stratified by Inflammatory Neuropathy Cause and Treatment (INCAT) disability scale scores. Previous IVIg treatment was discontinued after one final course ending the day before the first dose of fingolimod or placebo was given, whereas corticosteroids were tapered off over 8 weeks after randomisation. The primary endpoint was time to first confirmed worsening (≥1 point increase on the adjusted INCAT disability scale score versus baseline) and was assessed in the full analysis set, which consisted of all patients who underwent randomisation and had at least one efficacy assessment for the primary analysis. The survival distribution functions of time to first worsening were estimated within each treatment group according to the Kaplan-Meier survival distribution function and compared with a stratified log-rank test. The trial is registered with ClinicalTrials.gov, number NCT01625182. Of 106 participants randomly assigned between Jan 24, 2013, and March 10, 2016, 54 received fingolimod (41 who had been receiving IVIg and 13 who had been receiving corticosteroids) and 52 received placebo (41 who had been receiving IVIg and 11 who had been receiving corticosteroids). The trial ended for futility as recommended by an independent data monitoring committee after an interim analysis when 44 confirmed worsening events had occurred. At the end of the study, the survival estimate of the proportion of participants free from confirmed worsening was not significantly different between the fingolimod group (42%, 95% CI 23-60) and the placebo group (43%, 28-59; p=0·91). Adverse events occurred in 41 (76%) participants receiving fingolimod and 44 (85%) on placebo, and serious adverse events occurred in nine (17%) and four (8%) patients, respectively. The most common adverse events with fingolimod were headache (12 [22%] patients), hypertension (ten [19%]), and extremity pain (seven [13%]). Adverse events leading to study discontinuation occurred in seven (13%) participants on fingolimod and none on placebo. Fingolimod 0·5 mg once-daily was not better than placebo for the treatment of CIDP. Future trial designs should take account of the possibility that if IVIg is stopped abruptly, some patients might relapse soon afterwards whereas others might remain in remission. Novartis Pharma.

Post-Prostatectomy Radiotherapy in a Single Tertiary Institution: Outcomes Relating to Pre-Radiotherapy Prostate Specific Antigen

Objective: Timing of post-radical prostatectomy (RP) radiotherapy (RT) in patients with high risk prostate cancer continues to be debated. This is a retrospective review aiming to evaluate the influence of pre-RT prostate specific antigen (PSA) values on postprostatectomy RT outcomes in one Australian center. Method: Eligible patients were treated at the Royal Adelaide Hospital between January 2004 and December 2013, excluding those with nodal or distant metastatic disease pre-RT, or those who received neoadjuvant androgen deprivation therapy pre-RT. The primary endpoint of biochemical failure-free survival (bFFS) was defined as time from RP to date of biochemical failure (bF). Covariates of Gleason score, post-RP PSA, and pre-RT PSA were further analysed in relation to bFFS. Results: 103 of 122 patients underwent final analysis (8 were excluded for the above reasons; 11 had missing data). Median follow-up from RP was 60 months. Kaplan-Meier (KM) estimates of 1, 2, 3, 4 and 5-year survival probabilities were 93.5%, 83.4%, 82.4%, 76.6% and 71% respectively. There was no statistically significant correlation between bFFS and pathological T-stage (p=0.1), surgical margin involvement (p=0.7), or RT total dose (p=0.8). Analysis based on KM survival distributions and log-rank tests suggest that pathological Gleason score may have some influence on bFFS (p=0.04). Doubling the pre-RT PSA whilst holding all other factors and covariates constant, increases the hazard of bF at a particular time-point by approximately 19% on average. Conclusion: This single-institution retrospective study provides reasonable evidence for influence of pre-RT PSA on post-RP RT outcomes, arguing for earlier referral for RT.

Prognostic significance of neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and lymphocyte-to-monocyte ratio (LMR) in untreated and treated stage IV non-small cell lung cancer (NSCLC): An Australian cancer centre experience.

e20630 Background: Baseline high NLR predicts poorer outcome in stage IV NSCLC; post-treatment variation in NLR and potential impact on overall survival (OS) is not clear in this group. Prognostic implications of PLR and LMR are not well studied for stage IV NSCLC. We aimed to assess the prognostic role of NLR, LMR and PLR in stage IV NSCLC at diagnosis and post-treatment. Methods: A retrospective descriptive study of 279 patients with Stage IV NSCLC treated at our centre over five year period (2010 to 2015). NLR, PLR and LMR calculated at diagnosis and post first cycle chemotherapy/targeted treatment. Demographic variables summarized. Estimates of Kaplan-Meier survival distributions for OS were generated. Cox regression used to derive OS hazard ratios of predictive variables. Approval from ethics committee obtained. Results: Baseline NLR, PLR and LMR showed strong association with OS. A five-unit increase at baseline (treatment start date) in NLR was associated with an 11% increase in the hazard of death (HR = 1.115, 95% CI: 1.077-1.148) while the same increase in PLR was associated with 0.5% increase in the hazard of death (HR = 1.005, 95% CI:1-1.005). A five-unit increase in LMR at baseline was associated with a 50% reduction in the hazard of death (HR = 0.50, 95% CI:0.260-0.956). Post-treatment NLR, PLR and LMR were examined in 221 patients. Fitted models incorporating adjustment for baseline values demonstrated that increasing NLR correlated with a shorter OS (HR = 1.574, 95% CI:1.295-1.917) but no statistically significant relationship could be demonstrated for PLR nor LMR. Conclusions: While NLR, PLR and LMR showed significant associations with OS prior to treatment, only NLR showed an association with OS after treatment. OS was shorter for high NLR and high PLR at baseline compared to low NLR and low PLR at baseline. OS was longer for high LMR at baseline compared to low at baseline. An increase in NLR post-treatment correlated with a shorter OS. This is the first study examining the prognostic significance of all three ratios in patients with untreated and treated stage IV NSCLC.

Validation of the Killip-Kimball Classification and Late Mortality after Acute Myocardial Infarction

BackgroundThe classification or index of heart failure severity in patients with acute myocardial infarction (AMI) was proposed by Killip and Kimball aiming at assessing the risk of in-hospital death and the potential benefit of specific management of care provided in Coronary Care Units (CCU) during the decade of 60.ObjectiveTo validate the risk stratification of Killip classification in the long-term mortality and compare the prognostic value in patients with non-ST-segment elevation MI (NSTEMI) relative to patients with ST-segment elevation MI (STEMI), in the era of reperfusion and modern antithrombotic therapies.MethodsWe evaluated 1906 patients with documented AMI and admitted to the CCU, from 1995 to 2011, with a mean follow-up of 05 years to assess total mortality. Kaplan-Meier (KM) curves were developed for comparison between survival distributions according to Killip class and NSTEMI versus STEMI. Cox proportional regression models were developed to determine the independent association between Killip class and mortality, with sensitivity analyses based on type of AMI. Results: The proportions of deaths and the KM survival distributions were significantly different across Killip class >1 (p <0.001) and with a similar pattern between patients with NSTEMI and STEMI. Cox models identified the Killip classification as a significant, sustained, consistent predictor and independent of relevant covariables (Wald χ2 16.5 [p = 0.001], NSTEMI) and (Wald χ2 11.9 [p = 0.008], STEMI).ConclusionThe Killip and Kimball classification performs relevant prognostic role in mortality at mean follow-up of 05 years post-AMI, with a similar pattern between NSTEMI and STEMI patients.

Angioimmunoblastic T‐cell lymphoma with hyperplastic germinal centres (pattern 1) shows superior survival to patterns 2 and 3: a meta‐analysis of 56 cases

Angioimmunoblastic T-cell lymphoma (AITL) may present in patterns 1, 2 or 3, representing those with hyperplastic, regressed or effaced germinal centres (GCs), respectively, but the prognostic utility of this subclassification has not been previously validated. Twenty-five cases of AITL were reviewed immunohistologically and with in-situ hybridization for Epstein-Barr virus-encoded RNA and polymerase chain reaction for T-cell receptor gamma and immunoglobulin heavy chain clonality and followed for up to 120 months. Four cases had conventional hyperplastic GCs, two had floral GCs, and one had progressively transformed GCs, consistent with pattern 1 and one additional case had hyalinized GCs, consistent with pattern 2. The remaining 17 (pattern 3) cases lacked morphologically discernible GCs. The Kaplan-Meier survival distribution of pattern 1 cases (5-year survival 83%) was superior to that of pattern 2 and 3 cases [5-year-survival 36% (P = 0.0417)] only when combined with the 31 cases, seven of which were pattern 1, that Attygalle et al. had followed for up to 247 months and previously published. Furthermore, the development of B-lineage (classical Hodgkin or diffuse large-cell) lymphoma was associated exclusively with pattern 3 (P = 0.0057). Pattern 1 represents an indolent phase/grade of AITL, unassociated with the development of secondary B-lineage lymphoma and uninfluenced by treatment regimen.

Early planned institution of biventricular mechanical circulatory support results in improved outcomes compared with delayed conversion of a left ventricular assist device to a biventricular assist device

It is generally accepted that patients who require biventricular assist device support have poorer outcomes than those requiring isolated left ventricular assist device support. However, it is unknown how the timing of biventricular assist device insertion affects outcomes. We hypothesized that planned biventricular assist device insertion improves survival compared with delayed conversion of left ventricular assist device support to biventricular assist device support. We reviewed and compared outcomes of 266 patients undergoing left ventricular assist device or biventricular assist device placement at the University of Pennsylvania from April 1995 to June 2007. We subdivided patients receiving biventricular assist devices into planned biventricular assist device (P-BiVAD) and delayed biventricular assist device (D-BiVAD) groups based on the timing of right ventricular assist device insertion. We defined the D-BiVAD group as any failure of isolated left ventricular assist device support. Of 266 patients who received left ventricular assist devices, 99 (37%) required biventricular assist device support. We compared preoperative characteristics, successful bridging to transplantation, survival to hospital discharge, and Kaplan-Meier 1-year survival between the P-BiVAD (n = 71) and D-BiVAD (n = 28) groups. Preoperative comparison showed that patients who ultimately require biventricular support have similar preoperative status. Left ventricular assist device (n = 167) outcomes in all categories exceeded both P-BiVAD and D-BiVAD group outcomes. Furthermore, patients in the P-BiVAD group had superior survival to discharge than patients in the D-BiVAD group (51% vs 29%, P < .05). One-year and long-term Kaplan-Meier survival distribution confirmed this finding. There was also a trend toward improved bridging to transplantation in the P-BiVAD (n = 55) versus D-BiVAD (n = 22) groups (65% vs 45%, P = .10). When patients at high risk for failure of isolated left ventricular assist device support are identified, proceeding directly to biventricular assist device implantation is advised because early institution of biventricular support results in dramatic improvement in survival.

Prediction of Survival for Patients With Bullous Pemphigoid

To identify the prognostic factors of bullous pemphigoid (BP). Prospective study of patients with BP included in a randomized, controlled trial. Twenty dermatology departments in France. Patients One hundred seventy patients with BP initially treated with a 40-g/d dosage of clobetasol propionate cream (testing sample) and 171 patients initially treated with oral corticosteroids at a dosage of 0.5 or of 1.0 mg/kg per day, depending on the extent of BP (validation samples). The end point was overall survival during the first year after BP diagnosis. From the testing sample, associations of clinical and biological variables with overall survival were assessed using univariate and multivariate analyses. Selected predictors were included in a prognostic model. To verify that these predictors were not dependent on the treatment used, the model was then validated independently on the 2 series of BP patients treated with oral corticosteroids. Median age of the BP patients included in the testing sample was 83 years. The 1-year Kaplan-Meier survival rate was 74%. From univariate analysis, the main deleterious predictors were demographic factors (ie, older age and female sex), associated medical conditions (ie, cardiac insufficiency, history of stroke, and dementia), and low Karnofsky score, which is a measure of the patient's general condition. No factors directly related to BP, in particular extent of cutaneous lesions, were shown to be related to the patients' prognosis. From multivariate analysis, only older age (P = .02) and low Karnofsky score (P<.001) appeared independently predictive of death. From the Cox model including these 2 predictors, the predicted 1-year survival rates were 90% (95% confidence interval [CI], 85%-96%) for patients 83 years or younger with Karnofsky score greater than 40, 79% (95% CI, 69%-90%) for patients older than 83 years with Karnofsky score greater than 40, 65% (95% CI, 50%-86%) for patients 83 years or younger with Karnofsky score of 40 or less, and 38% (95% CI, 26%-57%) for patients older than 83 years with Karnofsky score of 40 or less. Kaplan-Meier survival distributions of patients from the validation samples appeared clearly separated according to these 4 categories and were in close agreement with corresponding predicted 1-year survival rates obtained from the testing sample. The prognosis of patients with BP is influenced by age and Karnofsky score. These predictors are easy to use and should facilitate the management of BP.

A population-based study of survival and discharge status for survivors after head injury.

Creation of a basis for the planning of rehabilitation after head injury in Denmark. Patients with cranial fractures or traumatic cerebral lesions occurring in Denmark in 1979-93 were identified by computerized searches in the national hospital register. Kaplan-Meier survival functions were calculated for these two categories. Hospital records for a random sample of 389 survivors in 1997 after cranial fracture, acute brain lesion or chronical subdural haematoma, which occurred in 1982, 1987 and 1992 in patients aged 15 years or more at injury, were reviewed. Survivors were characterized by age, gender, place and severity of injury, as well as neurophysical, speech and mental deficits at discharge from hospital. Acute/subacute mortality of hospitalized patients was 27% for cerebral lesions and 4% after cranial fracture. As attrition by death outweighed the decreasing incidence with time, the point prevalence of survivors in 1997 after brain lesions occurring in 1982, 1987 or 1992 was nearly the same, averaging 8.4 per 100 000 of the population above age 14. Half of them were severe, as defined by initial Glasgow Coma Score < 9 for more than 6 h. There was no tendency with time of injury with regard to percentage occurrence of neurophysical, speech or mental deficits at discharge. The calculated number of candidates for rehabilitation of personal activities of daily life briefly after injury was 9.8 per 100 000 with cerebral lesion and 1.2 per 100 000 with cranial fracture. The severest 1.5 per 100 000 per year were candidates for early intensive centralized rehabilitation in special centres. Five years after injury, < 0.13 per 100 000 population survived in the vegetative state. In this population-based study, survival after cranial fracture and cerebral lesion was described quantitatively through Kaplan-Meier survival distributions. Besides, patterns of severity, neurophysical and mental sequelae among survivors 5, 10 and 15 years post-injury were described. It was shown by examples how the study has been useful already for the planning of rehabilitation according to a national strategy from 1997 in Denmark. Besides, the results may serve as a reference for comparison with future studies.

Large sample properties of some survival estimators in heterogeneous samples

The Kaplan-Meier survival distribution estimator and Nelson hazard function estimator may not be appropriate descriptive devices for studies which involve nonhomogeneous populations. In this paper we investigate a mean survival estimator that incorporates important covariates associated with survival. The estimator is constructed based on Cox's regression model and a weighting according to any specified distribution of the baseline covariates. A mean hazard function estimator is also derived corresponding to the multiplicative intensity model which allows for recurrent outcome events. The weak convergence of each estimator is established using martingale and stochastic convergence results and its limiting variance is calculated. The strong and weak convergence of the Kaplan-Meier estimate and Nelson estimate are also explored in a nonhomogeneous population with an underlying proportional hazards model. It is shown that the Kaplan-Meier estimate remains a uniformly consistent estimate of S( t) = P( T > t) even in a heterogeneous sample if the censoring distribution is independent of covariates. However, when the censoring distribution depends on covariates, the Kaplan-Meier estimator can be quite biased even though the mean survival function estimator remains unbiased.