Article

Estimation of confidence limits for the cumulative probability of survival in life table analysis

Authors:
To read the full-text of this research, you can request a copy directly from the author.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the author.

... The following characteristics were studied: stage, type of surgery, persistence of residual tumour, nodal status, characteristics of peritoneal implants (noninvasive or invasive), stromal micro-invasion and treatment by adjuvant therapy (Table 1). Overall survival rates ( Table 2) were determined by using the Kaplan-Meier method, rates of evolutive invasive disease (Table 3) were determined by using 1-Kaplan-Meier and 95% confidence intervals were calculated by the Rothman method [8]. Exact confidence intervals were determined for the 0% or 100% survival rates in Tables 2 and 3. ...
... Prognostic factors for overall survival in 80 patients a Recurrence as borderline tumour and evolutive invasive disease. b 95% Rothman confidence interval[8]. c 95% Exact binomial confidence interval. ...
... Prognostic factors for evolution to invasive disease in 80 patients a 95% Rothman confidence interval[8]. b 95% Exact binomial confidence interval. CI, confidence interval. ...
Article
Background The aim of this study was to determine the prognostic factors for patients with advanced stage, low malignant potential ovarian tumour (LMPOT). Patients and methods A retrospective review of 80 patients with serous LMPOT and peritoneal implants treated at or referred to our institution was carried out. Results Sixty-five patients had non-invasive implants. Fifteen patients had invasive implants. Twenty-nine patients had stage II and 51 patients had stage III disease. Three patients died of evolutive invasive disease and four of complications of treatment. The only prognostic factor of progression to ‘evolutive invasive disease’ is the pathologic subtype of peritoneal implants. The 5-year rates of evolutive invasive disease in patients with non-invasive implants and invasive implants were 2% and 31%, respectively (P <0.002). Conclusions In this series, the only prognostic factor for patients with advanced stage borderline tumour is the type of peritoneal implant. More patients died of the treatment’s complications than of the disease itself. The patients’ prognosis with non-invasive implants seems to be excellent, and conservative management could be discussed in younger patients.
... 16 Candidate explanatory variables were tested for their association with time to DA-related death using univariable Cox proportional hazard regression models and Kaplan-Meier survival curves. 17,18 Median survival times were calculated as the smallest survival interval for which the Kaplan-Meier survival function was less than or equal to 0.5 with confidence intervals (CIs) estimated (where appropriate) using the method of Peto (1973) and Rothman (1978). 14,19,20 The assumption of proportionality of hazards was assessed both globally and, for each explanatory variable, using the weighted residual method. ...
... 17,18 Median survival times were calculated as the smallest survival interval for which the Kaplan-Meier survival function was less than or equal to 0.5 with confidence intervals (CIs) estimated (where appropriate) using the method of Peto (1973) and Rothman (1978). 14,19,20 The assumption of proportionality of hazards was assessed both globally and, for each explanatory variable, using the weighted residual method. 21 The proportional hazards assumption was further investigated for each explanatory variable by plotting the scaled Schoenfeld residuals for each dog as a function of DA diagnosis to death interval. ...
Article
Full-text available
Background: Disseminated aspergillosis (DA) in dogs has a guarded prognosis and there is a lack of a gold standard treatment protocol. Objective: To retrospectively assess survival times and factors influencing survival times. Animals: Dogs diagnosed with DA from January 2007 to June 2017. Methods: Disseminated aspergillosis case data were retrieved from 13 Australian veterinary referral centers, with a diagnosis confirmed with culture or PCR. Factors influencing survival time after diagnosis were quantified using a Cox proportional hazards regression model. Results: Thirty-four dogs met the study inclusion criteria. Twenty-two dogs were treated with antifungal treatment and 12 dogs received no antifungal treatment. Accounting for censoring of dogs that were either still alive on the date of data collection or were loss to follow-up, dogs treated with itraconazole alone (n = 8) had a median survival time (MST) of 63 (95% CI: 20-272) days compared to 830 (95% CI: 267-1259) days for the n = 14 dogs that received multimodal antifungal therapy ( χ 2 test statistic 8.6; df = 1; P < .01). The daily hazard of death (DHOD) for dogs with abnormally high serum creatinine concentration at the time of diagnosis was 7.4 (95% CI: 1.9-29) times that of dogs with serum creatinine within the reference interval. Conclusion and clinical importance: Serum creatinine concentration at the time of diagnosis is a useful prognostic indicator for survival after a diagnosis of DA. The MST for dogs treated with multimodal antifungal therapy is longer than itraconazole alone and warrant further investigation (P < .01).
... EFS was defined as the delay between ASCR and progression/relapse, second malignant neoplasm (SMN), or death from any cause, whichever occurred first. Survival curves were estimated using the Kaplan-Meier method [13] and 95% confidence intervals (95% CI) computed with the Rothman method [14]. Median follow-up was estimated using the Schemper method [15]. ...
Article
Full-text available
Intensive treatments including high-dose chemotherapy (HDC) with autologous stem cell rescue have improved high-risk neuroblastoma (HRNB) survival. We report the long-term health status of 145 HRNB survivors, alive and disease-free 5 years post HDC. Median follow-up was 15 years (range = 5–34). Six patients experienced late relapses, 11 developed second malignant neoplasms (SMNs), and 9 died. Event-free and overall survivals 20 years post HDC were 82% (95% CI = 70%–90%) and 89% (78%–95%), respectively. Compared with the French general population, the standardized mortality ratio was 19 (95% CI = 8.7–36.1; p < 0.0001) and the absolute excess risk was 37.6 (19.2–73.5). Late effects were observed in 135/145 patients (median = 3 events/patient); 103 had at least one severe event. SMNs arose at a median of 20 years post HDC and included carcinoma (n = 5), sarcoma (2), acute myeloid leukemia (2), melanoma (1), and malignant glioma (1). Non-oncologic health events included dental maldevelopment (60%), severe hearing loss (20% cumulative probability at 15 years), hepatic focal nodular hyperplasia (14%), thyroid (11%), cardiac (8%), and renal (7%) diseases and growth retardation (height-for-age z-score ≤ −2 for 21%). Gonadal insufficiency was near-universal after busulfan (40/43 females, 33/35 males). Severe late effects are frequent and progressive in HRNB survivors needing systematic very long-term follow-up.
... The crude cumulative incidence of local peritoneal relapse was calculated in a competing-risks framework. 16 Progression-free survival (PFS) was calculated from the day of definitive CRS to the date of the last follow-up visit, disease-progression, or death, whichever came first, according to the Kaplan-Meier method. The two-tailed log-rank test was used to assess the significance of the comparison between survival distributions. ...
Article
Full-text available
Background: Non-gynecologic rare peritoneal surface malignancies (PSMs) often are misdiagnosed as disseminated ovarian cancer and initially treated by gynecologic surgeons. This study aimed to assess whether these previous maneuvers (i.e., full surgical staging and/or cytoreductive attempts) affect outcomes after the definitive surgery performed in a tertiary center. Methods: The study reviewed 298 women affected by non-gynecologic PSM who underwent cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) after previous gynecologic surgery. Prior surgery was categorized as limited surgery (pLS: abdominal exploration with biopsy plus adnexectomy and/or appendectomy) or extended surgery (pES: full surgical staging or cytoreductive attempts including hysterectomy with bilateral salpingo-oophorectomy). Results: Of the 298 patients, 143 had pLS and 153 had pES. Morbidity was similar between the groups (P = 0.143), but the pES group had more severe urinary tract injuries (19 vs. 3; P < 0.001), longer operating time (585.9 vs. 506.7; P = 0.027), and more patients needing more than two anastomoses (41 vs. 26; P = 0.033). Age older than 55 years (odds ratio [OR] 2.42; P = 0.009) and number of anastomoses (OR 3.17; P = 0.002) correlated with severe morbidity; pES correlated with urinary tract grades 3 and 4 injuries (OR 7.9; P = 0.001). The 5-year cumulative incidence of locoregional relapse was significantly higher in the pES group (0.41 vs. 0.27; P = 0.012; median follow-up period, 69 months). The multivariate analysis identified a Peritoneal Carcinomatosis Index (PCI) higher than 20 and pES as independent risk factors. Conclusion: For women undergoing CRS±HIPEC for non-gynecologic PSM, the risk for locoregional relapse and severe postsurgical urinary tract complications is increased by pES. Therefore, prior full surgical staging or cytoreductive attempts without definitive gynecologic histology should be avoided. Prophylactic ureteral stenting and stricter oncologic follow-up assessment must be considered in this scenario.
... For example, Cutler and Ederer [20] proposed to use effective sample sizes which are corrected sample sizes according to the number of events observed. Different versions of this effective sample size were then discussed [17,21]. Finally, exact methods can be found for the estimation of the variance of the Kaplan-Meier estimator [22][23][24] which provide correct performances for small samples or heavy censorship. ...
Article
Full-text available
Median follow-up is often used to assess the quality of follow-up in longitudinal studies. Its relevance to evaluate the quality of the produced estimations seems obvious. However there is no clear decision rule on its use. We evaluated the quality of the Kaplan-Meier estimator of the survival rate according to diffeerent values of the median follow-up in order to propose an objective decision rule. We performed a sensitivity analysis via Monte Carlo simulations based on a reference sample, according to literature. The log(-log) confidence interval of the Kaplan-Meier estimator of survival rate was evaluated in terms of coverage probability and length. The Monte Carlo simulation was performed for 10,000 replications considering 170 values for the median follow-up. We could observe a high relationship between the quality of the estimation and the median follow-up. Heavy censorship leads to an overestimation of the survival rate and anti-conservative confidence intervals. From a threshold value of the median follow-up, the values of the criteria reach a plateau and good estimations of the survival rate are achieved. More precisely, simulations show that reliable Kaplan-Meier estimations cannot be obtained when the median follow-up is less than two thirds of the estimation time. This provides factual elements to evaluate results for long longitudinal studies or to assess ongoing cohorts.
... For example, id15 and id17 are seen to experience DLTs upon enrollment, and so undergo dose reductions; they exit at the end of periods 6 and 7, respectively, with determinations MTD 15 ∈ [0,1) and MTD 17 ∈ [1,2). An 80% confidence band is shown on the dose-survival curve, as constructed by R package km.ci (Strobl 2009) using a method due to (Rothman 1978). Thresholds corresponding to the bypass and stop rules are also marked with vertical lines. ...
Preprint
Full-text available
Background Coherence notions have a long history in statistics, as rhetorical devices that support the critical examination of statistical doctrines and practices. Within the special domain of dose-finding methodology, a widely-discussed coherence criterion has been advanced as a means to guard the conceptual integrity of formal dose-finding designs from ad hoc tinkering. This is not, however, the only possible coherence criterion relevant to dose finding. Indeed, a new coherence criterion emerges naturally when the near-universal practice of cohort-wise dose escalation is examined from a clinical perspective. Methods The practice of enrolling drug-naive patients into an escalation cohort is considered from a realistic perspective that acknowledges patients’ heterogeneity with respect to pharmacokinetics and pharmacodynamics. A new coherence criterion thereby emerges, requiring that an escalation dose be tried preferentially in participants who have already tolerated a lower dose, rather than in new enrollees who are drug-naive. The logical implications of this ‘precautionary coherence’ (PC) criterion are worked out in the setting of a 3+3 design. A ‘3+3/PC’ design that satisfies this criterion is described and visualized. A simulation study is performed, evaluating the long-run performance of this new design, relative to optimal 1-size-fits-all dosing. Results Under the PC criterion, the 3+3 dose-escalation design necessarily transmutes into a dose titration design. Two simple rules suffice to enable abandonment of low starting doses, and termination of escalation. The process of conducting the 3+3/PC trial itself models the application of a dose titration algorithm (DTA) that carries over readily into clinical care. The 3+3/PC trial also yields an interval-censored ‘dose-survival curve’ having a semantics that should prove familiar to oncology trialists. Simulated 3+3/PC trials yield DTAs over a median of 6 dose levels, achieving 50% improved population-level efficacy compared to optimal 1-size-fits-all dosing. Conclusions Dose individualization can be accomplished within a trial conducted along ‘algorithmic’ lines resembling those of the inveterate 3+3 design. The dose-survival curve arising from this ‘3+3/PC’ design has semantics that should prove familiar and conceptually accessible to oncology trialists, and also seems capable of supporting more formal statistical treatments of the design. In the presence of sufficient heterogeneity in individualized optimal dosing, a 3+3/PC trial outperforms any conceivable 1-size-fits-all dose-finding design. This fact eliminates the rationale for the latter designs, and should put an end to the further development and promulgation of 1-size-fits-all dose finding.
... 9 The prognostic analysis of EFS was done using a log-rank test for univariate analyses and Cox model for multivariate analysis. [10][11][12] The Cox model was stratified for the 3 studies. Hazard ratios from the multivariate Cox model were calculated. ...
Article
To identify the factors influencing outcome in childhood mature B-cell non-Hodgkin lymphoma and acute leukemia (B-NHL/AL) with central nervous system (CNS) disease (CNS+), we analyzed patients <18 years with newly diagnosed B-NHL/AL registered in 3 Lymphomes Malins B studies in France between 1989 to 2011. CNS+ was diagnosed on fulfillment of ≥1 of the following criteria: any L3 cerebrospinal fluid (CSF) blasts (CSF+), cranial nerve palsy, isolated intracerebral mass but also clinical spinal cord compression, and cranial or spinal parameningeal extension. Two hundred seventeen out of 1690 patients (12.8%) were CNS+. CNS+ was significantly associated with male gender, head/neck locations, Burkitt histology, high initial lactate dehydrogenase (LDH) level, and bone marrow involvement. CSF+ was the most frequent pattern of CNS+ (45%). For the 217 CNS+ patients, the 5-year event-free survival (EFS) and overall survival rates (95% confidence interval) were 81.5% (75.8% to 86.1%) and 83.9% (78.4% to 88.2%), respectively. In multivariate analysis, among CNS+ patients, low EFS was associated with CSF+, high initial LDH level, and poor response to cyclophosphamide, oncovin (vincristine), prednisone prephase. These findings have been considered for patient’s stratification in the international randomized phase 3 trial Inter-B-NHL-ritux 2010 for children and adolescents with high-risk B-NHL/AL with CNS+ CSF+ patients only receiving intensified chemotherapy.
... For example, Bickel and Doksum [50] reported a confidence interval based on the profile variance for the one-sample binomial proportion. Rothman [51] and Afifi, Elashoff and Lee JJ [52] described profile variance type of confidence intervals for survival probability. ...
Article
Full-text available
Background: Cohen's κ coefficient is often used as an index to measure the agreement of inter-rater determinations. However, κ varies greatly depending on the marginal distribution of the target population and overestimates the probability of agreement occurring by chance. To overcome these limitations, an alternative and more stable agreement coefficient was proposed, referred to as Gwet's AC1. When it is desired to combine results from multiple agreement studies, such as in a meta-analysis, or to perform stratified analysis with subject covariates that affect agreement, it is of interest to compare several agreement coefficients and present a common agreement index. A homogeneity test of κ was developed; however, there are no reports on homogeneity tests for AC1 or on an estimator of common AC1. In this article, a homogeneity score test for AC1 is therefore derived, in the case of two raters with binary outcomes from K independent strata and its performance is investigated. An estimation of the common AC1 between strata and its confidence intervals is also discussed. Methods: Two homogeneity tests are provided: a score test and a goodness-of-fit test. In this study, the confidence intervals are derived by asymptotic, Fisher's Z transformation and profile variance methods. Monte Carlo simulation studies were conducted to examine the validity of the proposed methods. An example using clinical data is also provided. Results: Type I error rates of the proposed score test were close to the nominal level when conducting simulations with small and moderate sample sizes. The confidence intervals based on Fisher's Z transformation and the profile variance method provided coverage levels close to nominal over a wide range of parameter combination. Conclusions: The method proposed in this study is considered to be useful for summarizing evaluations of consistency performed in multiple or stratified inter-rater agreement studies, for meta-analysis of reports from multiple groups and for stratified analysis.
... Complication rates for extended indications were compared using Fisher's exact test. Survivorship curves were constructed using the life-table method, with 95% confidence intervals calculated using Rothman's equation [24,25]. Separate endpoints were examined: revision for any reason and revision for confirmed loosening. ...
Article
Full-text available
Background: This prospective study evaluates outcomes and reoperation rates for unicompartmental knee arthroplasty (UKA) from a single non-designer surgeon using relatively extended criteria of degenerative changes of grade 2 or above in either or both non-operated compartments. Methods: 187 consecutive medial mobile bearing UKA implants were included after history, clinical assessment and radiological evaluation. 91 patients had extended clinical outcomes. Post-operative assessment included functional scoring with the Oxford Knee Score (OKS) and radiographic review. Survivorship curves were constructed using the life-table method, with 95% confidence intervals calculated using Rothman's equation. Separate endpoints were examined: revision for any reason and revision for confirmed loosening. Results: The mean follow-up was 3.5 years. The pre-operative OKS improved from a mean of 21.2 to 38.9 (Mann-Whitney U Test, p = < 0.001). Twelve Patients required further operations including 9 revisions. No patients developed deep infection and no surviving implants were loose radiographically. Survivorship at 7 years with endpoints of re-operation, revision and aseptic loosening at surgery or radiographically was 88.4% (95% CI 79.6-93.7), 93.1% (95% CI 85.5-96.9) and 97.3% (95% CI 91.2-99.2) respectively. The presence of pre-operative mild contralateral tibiofemoral or any extent of patellofemoral joint degeneration was of no consequence. Discussion: The indications for UKA are being expanded to include patients with greater deformity, more advanced disease in the patellofemoral joint and even certain features in the lateral compartment indicative of an anteromedial pattern of osteoarthritis (OA). However, much of the supporting literature remains available only from designer centres. This study represents a group of patients with what we believe to be wider indications, along with decisions to treat made on clinical grounds and radiographs alone. Conclusion: This study shows comparable clinical outcomes of UKA for extended indications from a high volume, high-usage non-designer unit.
... The Greenwood confidence interval has been replaced since by more accurate ones (as described in S2: 'Limitations of the Kaplan-Meier method') (Rothman 1978;Borkowf 2005;Miettinen 2008). But Kaplan's product-limit estimator still stands. ...
Article
In June 1958, Edward L Kaplan (1920–2006) and Paul Meier (1924–2011) published an innovative statistical method to estimate survival curves when including incomplete observations. The Kaplan–Meier (KM) method became the standard way of reporting patient survival in medical research. For example, the KM method is used in more than 70% of clinical oncology papers. With 44,319 Web of Science® citations as of November 2017, the report has become the most-cited statistics publication in the scientific literature. Part I of this report describes the KM method, its strengths and limitations, and the history and evolution of the method. In Part II we recount the biography of the remarkable mathematician Edward L Kaplan, PhD, and his unique contributions during the formulation of the KM method, as well as his contributions to science during his unique and productive career.
... 4,10,44 If these factors are constant in each series, life-table analysis and survivorship curves are useful in comparing different designs of prosthesis. 28,34,35 The correlation of survivorship with individual components enables the identifi cation of mechanisms of failure and allows selective improvement of components in a modular system. This led us to abandon metal-backed patellae in favour of a three-lug all-polyethylene component and to move from a fl at-on-fl at sagittal articulation to a more conforming design. ...
Article
A consecutive series of 235 total knee arthroplasties using the PFC system was followed prospectively for at least ten years in 186 patients. The operation was for osteoarthritis in 150 knees, for rheumatoid arthritis in 83, and for Paget’s disease and femoral osteonecrosis in one knee each. At the latest review 56 patients had died, fi ve were too ill to assess and three could not be traced. The PFC knee replacement utilised was a nonconforming posterior-cruciate-retaining prosthesis with a polyethylene insert which is fl at in the sagittal plane. The patella was resurfaced using a metal-backed component in 170 cases, but later in the series we used an all-polyethylene component in 22 knees; 43 patellae were not resurfaced. The survival without need for reoperation for any reason was 90% at ten years. Nineteen revisions were component-related due to failure of nine metal-backed patellae, nine polyethylene inserts, and one unresurfaced patella; two reoperations were for synovectomy (one for recurrent haemarthrosis and one for recurrent rheumatoid synovitis) and three were for metastatic joint infection. There were no revisions for aseptic loosening of femoral or tibial components, or the all-polyethylene patellar replacement. The PFC system provides good and predictable results in tricompartmental arthritis of the knee. Loosening appeared to be negligible, but there were wear-related problems in 8%. The change from a metal-backed patella and an increase in the contact area of the tibial insert should provide further improvement by minimising wear.
... Confidence limits were based on the 'effective number at risk' 3 and determined using Rothman's method. 4 ...
Article
We undertook a radiological analysis of 186 standard Souter implants to determine survivorship and to analyse the pattern of failure in those needing revision. The implants had been inserted as a primary procedure in patients with rheumatoid arthritis of the elbow at our hospital over the last 12 years. Taking revision as an endpoint, the survivorship after 12 years was 87%. If, however, revision and loosening, defined as the Hindex value equivalent to demarcation of 1 mm around the whole implant, are also included, the survivorship falls to 80%. Of the 24 implants revised, 18 (75%) were for problems with the humeral component, three (12.5%) with the ulnar component and three (12.5%) for instability. Loosening of the humeral component occurred when the implant extended into the humerus, with the tip moving anteriorly on to the anterior humeral cortex. Our study indicates that loosening can be predicted by the rate of change in this angle of extension of the prosthesis.
... Survival rates of the uncemented Thompson's prosthesis were calculated as described by Murray et al., with revision of the implant being used as the endpoint in this study [9]. 95% confidence intervals were calculated using the method described by Rothman [10]. The survival curve was produced using Stata Version 13.1 (StataCorp, Texas, USA). ...
Article
Full-text available
Background: Guidelines on the management of displaced intracapsular fractures recommend using an Orthopaedic Data Evaluation Panel-rated cemented implant. Prior to the National Institute for Health and Care Excellence guidelines, uncemented implants were commonly used in the UK. Methods: We retrospectively examined the outcomes of patients with uncemented Thompson's hemiarthroplasties at our unit, between April 2005 and December 2010. Patients who underwent revision surgery before December 2011 were identified. Implant survival calculation utilised the primary outcome of revision to total hip arthroplasty, revision hemiarthroplasty or excision arthroplasty. Patients who died post-operatively were identified and censored. Results: A total of 1445 patients received uncemented Thompson's implant. Patient mean age was 82 years with 76% female. Forty-six (3.2%) patients required revision with 15% performed within 30 days of surgery and 62% within 1 year. Reasons for revision were infection (0.83%), acetabular erosion (0.83%) and loosening (0.62%). Twenty-seven patients (59% of total revisions) underwent revision to THA, 14 (30%) to excision arthroplasty and 5 (11%) to revision hemiarthroplasty. Cumulative survival rate was 98% at 1 year and 95% at 5 years. Thirty-day mortality was 7.1%. One-year mortality was 28.1%. Conclusion: Current guidelines strongly favour cemented hemiarthroplasty. Recognition that fractured hip patients are a non-homogeneous group is important. In patients with limited life expectancy, an uncemented Thompson is a quick, simple, palliative solution to early mobilisation. Correct surgical technique avoids using cement in this cohort, which is most vulnerable to bone cement implantation syndrome. Cost-effective resource utilisation with an increasingly elderly population remains a surgical responsibility.
... Wilson quadratic 95% confidence intervals were calculated using the method described by Rothman. (38) ...
Article
We prospectively studied 152 hips in 126 patients with a proximally porous and hydroxyapatite coated double tapered straight femoral prosthesis. The follow-up period was 4.2 to 9.6 years. All but two prostheses showed osseo-integration. The mean Harris hip score (HHS) increased from 40 to 92 points and 97% of the patients had no or only mild pain. Limping and the use of walking aids decreased substantially after the operation. The morphology of positive bone remodelling with either endosteal bone formation or periosteal bone formation was correlated with the stem fit in the medullary canal (p=0.0001). Negative bone remodelling features could not be demonstrated. There was no correlation between the clinical parameters and the radiological phenomena.
... A life table was constructed and a survival analysis was performed to calculate the cumulative survival rate (12,13). Confidence intervals were calculated according to the equation of Rothman (14). Revision was defined as the endpoint. ...
Article
The first consecutive series of 250 implantations of an uncemented, proximally anchored straight femoral shaft prosthesis in 236 patients is included in this prospective follow-up study. Average time of follow-up evaluation was 11.8 years (range 10.6 - 13.7 years). At follow-up, 41 patients with 44 hips had died and three could not be located. Eight hips were revised, two for infection, one for recurrent dislocation, two for rapid subsidence due to component undersizing and one for aseptic loosening of a varus-malaligned stem; two radiologically well-fixed stems were revised during acetabular revision. Survival estimate showed an overall survival of 96.6% after 14 years (confidence limits: 98.4% (upper) and 92.8% (lower)). The median Harris hip score at time of follow-up was 81.4 points; clinical results were compromised by high loosening rates (68.8%) of the threaded cup, which had been used in 165 of the 250 cases. Radiologically, tiny reactive lines (< 2mm) were frequently present in the distal zones of the femoral shaft, but could be found in the proximal anchoring zone in only one case. In conclusion, the long-term follow-up results with this femoral component are encouraging and are comparable to those of modern techniques of cementing in primary total hip arthroplasty.
... One can calculate the confidence interval of survival rates to assess the variability in the same (Greenwood, 1926;Rothman, 1978;Anderson et al, 1982). ...
Thesis
Full-text available
Knowledge of survival is essential in the community level management of a disease. Broadly, there are two approaches of population-based study of survival from a disease, the direct (i.e., classical) approach and the indirect approach. With classical approaches, survival studies deal with evaluating overall performance of a group of patients in terms of quality and quantity of life after diagnosis/treatment. There are numerous difficulties in the conduct of a population-based survival study in the context of developing countries, including India. While planning a population based survival study, one has to consider the possibility of a substantial amount of financial and other resources including the time required. Subsequently, loss to follow-up is a typical problem encountered in survival studies, causing biased estimates. In view of these difficulties with the classical approach, the overall aim of the present study was to propose an indirect methodology for the study of survival. Specific objectives were to a) suggest an indirect methodology for the study of survival, b) demonstrate empirical application of the methodology and c) validate the proposed methodology. Proposed methodology is based on life table techniques and uses current data on incidence and mortality from the disease. It involves the estimation of expected years free of disease (EYFD), expected years with disease (EYWD), expected years of life lost (EYLL) and average duration of disease (ADD) and their comparison over a time period. Empirical application was carried out for mouth and lung cancers in males and cancers of breast and cervix in females as well as for all sites combined together in each sex. Cancer incidence and mortality data by age and sex for the years 1989, 1993, 1997 and 2001 were obtained from published reports of Mumbai Cancer Registry, India. All causes of deaths for these years were obtained from Mumbai Municipal Corporation. Three life tables were constructed by applying various attrition factors: (a) risk of death from all causes; (b) risk of incidence and that of death from other causes; and (c) risk of death from other causes only. The expectation of life from the second life table gave EYFD. EYWD and EYLL were calculated by suitable subtractions among three expectations of life. ADD was calculated by dividing person years lived with disease by number developing the disease. In order to arrive at total number developing the disease, a break-up of those arriving to the last open-end age interval in second life table was required. The break-up was required into those developing and those not developing the disease before dying. This was obtained by employing an iteration procedure. It was noted that during 1993-2001, EYFD for all sites increased from 59.4 to 62.1 and from 63.8 to 66 years in males and females respectively. EYLL was about 0.8 year in males and 1 year in females. Similarly, EYWD was 0.6 and 1 year in males and females respectively. ADD for all sites varied from 4 to 4.7 years in both sexes. It was about 6 years for mouth cancers and 2 years for lung cancers in males and 4-5 years for breast and cervical cancers in females. An indication of improvement in survival indices was observed in females. Given the difficulties in conduct of classical survival studies, the proposed method may provide a useful tool for the study of survival, especially for ’all sites’ and major sites of cancer, viable for the developing countries. Proposed method may also be useful in having a regular audit of the prognostic factors prevailing in a population. Further, it also has potentials to be utilized for the estimation of various indices of burden of disease, like, PYLL, DALY, etc.
... where i is the time interval and n is the number of patients at risk in the time interval i. 208,210 Accordingly, the confidence limits (CL) are calculated according to the following formula 209 when M is an effective number at risk and P is cumulative survival at the given time interval (expressed as proportion and not as percentage). This mathematical expression is based on the theoretical assumption presented by Rothman 212 and popularized by Murray et al. 208 The mathematical basis of these assumptions will not be discussed in this presentation, which is more of a practical nature. The interested reader is referred to these extensive statistical reports that are given in the "References" section. ...
... We adjusted Greenwood's formula accordingly to calculate variance, and we calculated 95% CIs for a binomial proportion with the same value and variance. 18 Curves are plotted until the 75th percentile of the distribution of observation times. All p values are two-sided. ...
... Crude cumulative incidence of peritoneal and distant metastases were calculated in a competing risks framework. 21 In peritoneal (distant) recurrence analysis, deaths without evidence of disease, and distant (peritoneal) metastasis, whichever occurred first, were regarded as competing events. OS and PFS were dated from the day of surgery to the date of death from any cause, or first recurrence involving any site, respectively. ...
Article
Full-text available
Background Cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (HIPEC) are maximally effective in early-stage colorectal cancer peritoneal metastases (CRC-PM); however, the use of HIPEC to treat subclinical-stage PM remains controversial. This prospective two-center study assessed adjuvant HIPEC in CRC patients at high risk for metachronous PM (www. clinicaltrials. gov NCT02575859). Methods During 2006–2012, a total of 22 patients without systemic metastases were prospectively enrolled to receive HIPEC simultaneously with curative surgery, plus adjuvant systemic chemotherapy (oxaliplatin/irinotecan-containing ± biologics), based on primary tumor-associated criteria: resected synchronous ovarian (n = 2) or minimal peritoneal (n = 6) metastases, primaries directly invading other organs (n = 4) or penetrating the visceral peritoneum (n = 10). A control group retrospectively included 44 matched (1:2) patients undergoing standard treatments and no HIPEC during the same period. The cumulative PM incidence was calculated in a competing-risks framework. Results Patient characteristics were comparable for all groups. Median follow-up was 65.2 months [95 % confidence interval (CI) 50.9–79.5] in the HIPEC group and 34.5 months (95 % CI 21.1–47.9) in the control group. The 5-year cumulative PM incidence was 9.3 % in the HIPEC group and 42.5 % in the control group (p = 0.004). Kaplan–Meier estimated 5-year overall survival (OS) was 81.3 % in the HIPEC group versus 70.0 % in the control group (p = 0.047). No operative death occurred. Grade 3–4 [National Cancer Institute Common Terminology Criteria for Adverse Events (NCI–CTCAE) version 4] morbidity rates were 18.2 % in the HIPEC group and 25 % in controls (p = 0.75). At multivariate analysis, HIPEC correlated to lower PM cumulative incidence [hazard ratio (HR) 0.04, 95 % CI 0.01–0.31; p = 0.002], and better OS (HR 0.25, 95 % CI 0.07–0.89; p = 0.039) and progression-free survival (HR 0.31, 95 % CI 0.11–0.85; p = 0.028). Conclusion Adjuvant HIPEC may benefit CRC patients at high-risk for peritoneal failure. These results warrant confirmation in phase III trials.
... The survival probabilities were estimated according to the Kaplan-Meier method (37). The curves carry Rothman's 95% confidence intervals (42). Survival curves were compared by the logrank test (36). ...
Article
Purpose: Our Phase II trial using bleomycin, epirubicin, and cisplatin (BEG) protocol in the treatment of locoregionally advanced undifferentiated nasopharyngeal carcinoma (UCNT) patients has shown encouraging results with high objective response, disease-free survival, and overall survival rates. To establish the value of this BEC regimen as neoadjuvant chemotherapy, we initiated in 1989 a large international Phase III trial. It compares three cycles of EEC followed by radiotherapy to radiotherapy alone. Methods and Materials: From November 1989 to October 1993, 339 patients with negative metastases workup, stratified by accrual center have been randomized, 168 to radiotherapy alone and 171 to chemotherapy plus radiotherapy, All patients characteristics were well balanced in both arms. There was a quality control/data verification by specialist panel (radiology, histology, radiotherapy, chemotherapy) and external policy board expert every 60-80 patients having completed treatment. Results: With a median follow-up of 49 months (range: 23-70), despite an excess of treatment-related deaths in the neoadjuvant chemotherapy arm (8 vs. 1%), there is a significant difference in disease free survival favoring the chemotherapy arm (p < 0.01). The proportion of local and/or regional metastases was comparable in both arms. No difference in overall survival is seen but the numbers of events needed for analysis has not yet been reached. Conclusions: BEC type neoadjuvant chemotherapy has a significant impact in the natural history of UCNT. Further follow-up is needed to establish an eventual overall survival difference.
... Journal editorial policies in medicine and psychology encourage researchers to supplement significance testing by reporting confidence intervals (CIs) as well as effect size (ES) statistics. As Fidler, Thomason, Cumming, Finch, and Leeman (2004) note, this movement started in medicine as early as the 1980s (see Rothman 1975Rothman , 1978aRothman , 1978b. In psychology, in the past 15 years or so, there has been renewed emphasis on reporting ESs because of editorial policies requiring ESs (e.g., Murphy, 1997;Thompson, 1994) and official support for the practice. ...
Article
Full-text available
The probability coverage of intervals involving robust estimates of effect size based on seven procedures was compared for asymmetrically trimming data in an independent two-groups design, and a method that symmetrically trims the data. Four conditions were varied: (a) percentage of trimming, (b) type of nonnormal population distribution, (c) population effect size, and (d) sample size. Results indicated that coverage probabilities were generally well controlled under the conditions of nonnormality. The symmetric trimming method provided excellent probability coverage. Recommendations are provided.
... A combination of the information provided by the persistence of HIV-1 antibodies at 15 months of age in children born to HIV-1 seropositive mothers and by the excess mortality in this group compared with the reference cohort constituted of children born to HIV-l seronegative mothers (Method 1). Method 1 was originally proposed by Halsey et al. (14) in Haiti and was subsequently revised (1 8) to obtain the following equation: (26). ...
Article
Full-text available
The authors report the results of the first 2 years of follow–up of a prospective cohort study on the mother–to–child transmission of human immunodeficiency virus type 1 (HIV–1) and its determinants which started in November 1988 in Kigali, Rwanda. The study sample consists of 218 newborns of 215 HIV–1 seropositive women matched to 218 newborns of 216 HIV–1 seronegative women of the same age and parity. They were followed every 2 weeks during the first 2 years of follow–up. HIV–1 antibodies were detected by enzyme–linked immunoadsorbent assay and Western blot at 3–month intervals. Two methods of calculating the mother–to–child transmission rate were used: method 1 combines the information provided by the persistence of HIV–1 antibodies at 15 months of age in children born to HIV seropositive mothers and the excess mortality in this group compared with the cohort of children born to HIV seronegative mothers; method 2 is a case–by–case evaluation of all the children born to HIV seropositive mothers. A logistic regression model was used to study the determinants of transmission. The probability of survival at 24 months of age was 81% (95% confidence interval (Cl) 75–86) in children born to seropositive mothers, compared with 95% (95% Cl 92–98) in children born to seronegative mothers (p < 0.001). The mother–to–child transmission rate calculated with method 1 was 25.7% (95% Cl 18.8–32.5). With method 2, the medium estimate was 24.7%. In the multivariate analysis, a CD4/CD8 ratio <0.5 was the only maternal factor statistically associated with an increased risk of mother–to–child transmission of HIV–1 (odds ratio=2.9, 95% Cl 1.2–7.2). The authors' findings present evidence for a higher mother–to–child transmission rate of HIV–1 in children born in Rwanda than in industrialized countries. Am J Epidemiol 1993;137:589–99.
Article
Full-text available
Primary Mediastinal large B-cell Lymphoma (PMLBL) is a rare entity predominantly affecting adolescents and young adults. Recently, an international phase 2 trial in pediatric patients using dose-adjusted etoposide, doxorubicin, and cyclophosphamide with vincristine and prednisone plus rituximab (DA-EPOCH-R) failed to reproduce excellent survival reported in some adult studies. The optimal therapy regimen needs to be determined in this disease. French prospective LMB2001 trial included all patients < 18 years with mature B-cell lymphoma treated in French centers. For patients with PMLBL, treatment included 4 to 8 courses of Lymphomes Malins B (LMB)-based chemotherapy without radiotherapy. From 2008, rituximab was added before each chemotherapy course. From 09/2001 to 03/2012, 42 patients with PMLBL were registered. The median age was 15 years (range 8-18). 21 patients were treated by chemotherapy plus rituximab. The median follow-up was 7.1 years (Interquartile range, 5.8-11.1). Five-year event-free (EFS) and overall survival (OS) were 88.1% (95%Confidence Interval (CI), 75.0%-94.8%) and 95.2% (95%CI, 84.0%-98.7%) for the whole population. The 5-year EFS was 81.0% (95%CI, 60.0%-92.3%) and 95.2% (95%CI, 77.3%-99.2%) (HR=0.24 (95%CI 0.03-2.2)) and 5-year OS was 90.5% (95%CI, 71.1%-97.3%) and 100% for patients treated without and with rituximab, respectively. Only 1/21 patients treated with rituximab and LMB-based chemotherapy had local early treatment failure but achieved prolonged complete remission with second-line chemotherapy and radiotherapy. Intensive LMB-based chemotherapy with rituximab achieved excellent survival in children/adolescents with PMLBL. Further international prospective studies are required to confirm these results in this population.
Article
Background and aimsPrognosis of hepatoblastoma patients has increased with cisplatin-based chemotherapy and high-quality resection including liver transplant. Consequently current risk-adapted therapeutic strategy aims to reduce long-term side effects in patients with standard risk disease.Methods We report long-term mortality and morbidity data concerning 151 2-year hepatoblastoma survivors treated with SIOPEL risk-adapted strategies (sex-ratio M/F = 1.6, median age at diagnosis = 2.6 years [range 0–17.7], median year at diagnosis = 2008 [1994–2017]). Fifty-three patients had loco-regional risk factors VPEFR, 12 were PRETEXT-IV and 30 were metastatic. All received cisplatin and 84 anthracyclines. Twelve had liver transplant. To assess hearing, renal and cardiac functions, audiograms were performed in 116/151 patients (76.8%), glomerular filtration rate in 113/151 (74.8%) and cardiac ultrasound in 65/84 (77.4%) anthracycline-exposed patients.ResultsWith a median follow-up of 9.4 years (range 2.1–25.8), four late relapses, one second malignancy (Acute Myeloid Leukemia AML-M5) and two deaths (one from hepatoblastoma, one from AML) occurred. The 10-years event free survival and overall survival probabilities were 95.5% (95% CI 91.9–99.1) and 98.7% (95% CI 96.8–100), respectively. Sixty-eight non-oncologic health-events included 57 cases of hearing loss (including 25 Brock 3–4), three liver cirrhosis, three pre-operative portal cavernoma, two focal nodular hyperplasia, two grade-1 chronic kidney diseases and one asymptomatic cardiac dysfunction were reported. Ototoxicity was significantly associated with cisplatin cumulative dose (OR = 2.07, 95% CI 1.32–3.24, p = 0.001) and carboplatin exposure (OR = 3.14, 95% CI 1.30–7.58, p = 0.01) in multivariable analysis adjusted for sex and age at diagnosis.Conclusions With current risk-adapted strategies, hepatoblastoma is a highly curable disease, with very rare relapses, and few late effects except hearing loss which remains a serious condition in these very young patients.
Article
Background: Both the direct anterior approach (DAA) and posterior approach (PA) to THA have known advantages and disadvantages. The comparison between DAA and PA THA has been widely explored during the early postoperative period. However, few randomized trials have compared these approaches at a minimum follow-up of 5 years; doing so would be important to establish any differences in mid-term outcomes or complications. Questions/purposes: We performed a randomized trial comparing DAA and PA in THA in terms of (1) patient-reported outcome scores, (2) quality of life and functional outcomes assessed by the EQ-5D and 10-meter walk test results, (3) radiographic analysis, and (4) survivorship and surgical complications at a minimum of 5 years follow-up. Methods: Two hip specialist surgeons performed both DAA and PA THA using the same THA components at two hospital sites. One hundred twelve patients on the elective THA surgical waitlist were invited to participate in the study. Thirty-four patients did not meet the study's inclusion criteria and were excluded, and three patients declined to participate in the study. The remaining 75 patients who were eligible were randomized into DAA and PA groups. Thirty-seven patients were initially randomized to receive DAA THA, but two did not and were excluded, resulting in 48% (35 of 73) of patients who received DAA THA; 52% (38 of 73) of patients were randomized into and received PA THA. Over a minimum 5 years of follow-up, 3% (1 of 35) of DAA patients were lost to follow-up, and none of the patients undergoing PA THA were lost. A per-protocol analysis was adopted, resulting in further patients being excluded from analysis. Of the 73 study patients, 99% (72; DAA: 35, PA: 37) were analyzed at 1 year, 95% (69; DAA: 34, PA: 35) were analyzed at 2 years, and 72% (52; DAA: 23, PA: 29) were analyzed at 5 years. The primary outcome was the Oxford Hip Score (OHS) and WOMAC score. Secondary outcomes included the EQ-5D and EQ-5D VAS scores, 10-meter walk test results, radiographic evidence of loosening (femoral: lucency > 2 mm at the implant-bone interface, subsidence > 2 mm; acetabular: migration or change in inclination), 5-year survivorship analysis from all-cause revisions, and surgical complications. The study was powered to detect a 10-point difference in the WOMAC score, which is equivalent to the minimum clinically important difference (MCID). Results: There were no differences in primary outcomes (OHS and WOMAC scores) or secondary outcomes (EQ-5D scores, EQ-5D VAS scores, and 10-meter walk test result) between the DAA and PA groups at the 5-year follow-up interval. The median (range) OHS at 5 years was 46 (16 to 48) for DAA and 47 (18 to 48) for PA groups (p = 0.93), and the median WOMAC score was 6 (0 to 81) for DAA and 7 (0 to 59) for PA groups (p = 0.96). The median EQ-5D score was 1 (0.1 to 1) for DAA and 1 (0.5 to 1) for PA groups (p = 0.45), and the median EQ-5D VAS score was 85 (60 to 100) for DAA and 95 (70 to 100) for PA groups (p = 0.29). There were no cases of component loosening on radiographs. There was no difference in component survival between the two approaches at 5 years (DAA: 97% [95% CI 85% to 100%] versus PA: 97% [95% CI 87% to 100%]). Eight of 23 patients in the DAA group reported decreased sensation in the lateral femoral cutaneous nerve distribution. Conclusion: DAA and PA are both effective approaches in performing primary THA. Each approach has its associated risks and complications. The choice of THA should be based on individual patient factors, surgeon experience, and shared decision-making. Early registry data indicate DAA and PA THA are comparable, but longer-term data with larger numbers of patients will be required before one can safely conclude equal survivorship between both approaches. Level of evidence: Level I, therapeutic study.
Preprint
Full-text available
Background The confidence interval for survival probability at a fixed time point provides valuable information on how the subject performs in terms of survival rate. However, in a two-arm trial when the sample size in each group is small or when the distribution of events that occurred within the group is skewed, the confidence interval might become very unstable, and thus may not provide accurate information for estimating survival rate. In addition, when there are other covariates available in the dataset, it is important to select those significant variables and include them in the model. On the other hand, researchers such as physicians who pay more attention to the final result often analyze the treatment group and control group separately, which may lead to inaccurate prediction. Methods In this study, two treatment groups are combined, and the group indicator variable is considered as a covariate and is included in the model for computation. Yuan and Rai’s adjusted effective sample size methods are further extended along with Cox proportional hazard model, Weibull model, and log-logistic model to compute predicted fixed-term overall survival probabilities and corresponding confidence intervals with other covariates adjusted. Simulations are conducted to obtain coverage probability. Results In a single model, Wilson-Peto provides better confidence intervals than Kaplan-Meier, especially in the middle and later stages. In addition, AC-Peto produces better coverage probability at all time points. In a multivariate model, the log-logistic method provides both better confidence intervals and coverage probability than Cox regression model at all stages. Conclusions This paper provides a guideline on how one should correctly analyze survival data with the most appropriate method. Depending on the dataset, it is important to consider methods other than traditional Kaplan-Meier and Cox regression models when evaluating survival outcomes.
Article
Relapses involving the central nervous system (CNS) are rare in children and adolescents with ALK+ anaplastic large cell lymphoma (ALCL) treated with regimens including CNS prophylaxis. Early identification of patients at high‐risk for CNS relapse would enable stratification and better adaptation of initial treatment especially in the light of the upcoming targeted therapies with limited CNS penetration. We analyzed clinical and histological data of all ALK+ALCL patients with CNS relapse registered in ALCL99‐database with the aim to describe risk factors and outcome. Characteristics of patients with no relapse, relapse without CNS involvement and CNS relapse were compared. At a median follow‐up of 8 years (0.05–18 years), a CNS involvement was reported at first or subsequent relapse in 26/618 patients. Median interval between initial diagnosis and first CNS relapse was 8 months (IQR 5.55–10.61/range 1.31–130.69). The 5‐year cumulative risk of CNS relapse was 4% (95% CI 2.9–5.5). Bone marrow involvement, peripheral blasts and CNS involvement at diagnosis were more frequent in patients with CNS relapse than in patients with no relapse or with relapse with no CNS involvement. The treatment of CNS relapse was heterogeneous. The median survival after CNS relapse was 23.7 months. Eleven patients were alive at last follow‐up. Three‐year overall survival after CNS relapse was 48.70% (95% CI 30.52–67.23).
Article
Objective: To describe characteristics and outcome of pediatric ovarian immature teratomas (IT) to better define the place of chemotherapy. Methods: Children with ovarian IT enrolled in TGM95 and TGM2013 studies were analyzed. Norris grading and International Federation of Gynecology and Obstetrics staging system were used. Results: Thirty-six cases were identified with a median age of 11 years (range = 1-18): 35 of 36 stage I (17 stage IA, 13 stage IC, and 5 stage IX), including seven patients with gliomatosis peritonei (GP), and 1 stage IIIB (IT peritoneal implants). Centrally reviewed Norris grading was performed in 31 cases: 14 grade I and 17 grade II/III tumors. All patients underwent upfront surgery: 19 unilateral oophorectomy, 14 unilateral adnexectomy, 2 unilateral cystectomy, and 1 bilateral cystectomy. No extensive GP surgery was performed. Six patients received adjuvant vinblastin, bleomycin, and cisplatinum because of tumor rupture (n = 5, including two patients with GP) or stage III (n = 1). After a median follow-up of 39.5 months (range = 6-238), two events occurred 10 and 11 months after diagnosis: one bilateralization (initial stage IX, grade I) and one IT peritoneal relapse (initial stage IA, grade II), respectively. Both were successfully rescued by platinum-based chemotherapy and delayed surgery. No stage IC patients treated without adjuvant chemotherapy relapsed (four grade I and three grade III). None of the seven patients with GP progressed. Five-year event-free survival and overall survival were 94% (95% CI = 81-98%) and 100%. Conclusions: The current series confirms the excellent prognosis of pediatric ovarian IT, arguing for conservative surgical approach in GP and against systematic adjuvant chemotherapy, even in ruptured tumors.
Article
Full-text available
Purpose To investigate the effect of adding concurrent chemotherapy (CT) to cetuximab plus radiotherapy (RT; CT-cetux-RT) compared with cetuximab plus RT (cetux-RT) in locally advanced squamous cell carcinoma of the head and neck (LA-SCCHN). Patients and Methods In this phase III randomized trial, patients with N0-2b, nonoperated, stage III or IV (nonmetastatic) LA-SCCHN were enrolled. Patients received once-daily RT up to 70 Gy with weekly cetuximab or with weekly cetuximab and concurrent carboplatin and fluorouracil (three cycles). To detect a hazard ratio (HR) of 0.64 for progression-free survival (PFS) with 85% power at a two-sided significance level of P = .05, 203 patients needed to be included in each arm. Results Four hundred six patients were randomly assigned to either CT-cetux-RT or cetux-RT. Patient and tumor characteristics were well balanced between arms, including p16 status. With a median follow-up of 4.4 years, the HR for PFS favored the CT-cetux-RT arm (HR, 0.73; 95% CI, 0.57 to 0.94; P = .015), with 3-year PFS rates of 52.3% and 40.5% and median PFS times of 37.9 and 22.4 months in the CT-cetux-RT and cetux-RT arms, respectively. The HR for locoregional control was 0.54 (95% CI, 0.38 to 0.76; P < .001) in favor of CT-cetux-RT. These benefits were observed regardless of p16 status for oropharynx carcinomas. Overall survival (HR, 0.80; P = .11) and distant metastases rates (HR, 1.19; P = .50) were not significantly different between the two arms. The CT-cetux-RT arm, compared with cetux-RT, had a higher incidence of grade 3 or 4 mucositis (73% v 61%, respectively; P = .014) and of hospitalizations for toxicity (42% v 22%, respectively; P < .001). Conclusion The addition of concurrent carboplatin and fluorouracil to cetux-RT improved PFS and locoregional control, with a nonsignificant gain in survival. To our knowledge, this is the first evidence of a clinical benefit for treatment intensification using cetux-RT as a backbone in LA-SCCHN.
Article
Background: We report 15-year survival, clinical, and radiographic follow-up data for the Press-Fit Condylar Sigma total knee arthroplasty. Methods: Between October 1998 and October 1999, 235 consecutive TKAs were performed in 203 patients. Patients were reviewed at a specialist nurse-led clinic before surgery and at 5, 8-10, and 15 years postoperatively. Clinical outcomes, including Knee Society Score, were recorded prospectively at each clinic visit, and radiographs were obtained. Results: Of our initial cohort, 99 patients (118 knees) were alive at 15 years, and 31 patients (34 knees) were lost to follow-up. Thirteen knees (5.5%) were revised; 5 (2.1%) for infection, 7 (3%) for instability, and 1 (0.4%) for aseptic loosening. Cumulative survival with the end point of revision for any reason was 92.3% at 15 years and with revision for aseptic failure as the end point was 94.4%. The mean Knee Society Score knee score was 77.4 (33-99) at 15 years, compared with 31.7 (2-62) preoperatively. Of 71 surviving knees for which X-rays were available, 12 (16.9%) had radiolucent lines and 1 (1.4%) demonstrated clear radiographic evidence of loosening. Conclusion: The Press-Fit Condylar Sigma total knee arthroplasty represents a durable, effective option for patients undergoing knee arthroplasty, with excellent survival and good clinical and radiographic outcomes at 15 years.
Article
We describe our experience with the implantation of 325 Exeter Universal hips. The fate of every implant was known. The procedures were undertaken by surgeons of widely differing experience. At follow-up at 12 years, survivorship with revision of the femoral component for aseptic loosening as the endpoint was 100% (95% CI 98 to 100). Survivorship with revision of the acetabular component for aseptic loosening as the endpoint was 96.86% (95% CI 93.1 to 98.9) and that with any reoperation as the endpoint 91.74% (95% CI 87.7 to 95.8). No adverse features have emerged as a consequence of the modular connection between the head and neck of the implant.
Article
We retrospectively assess 171 consecutive total hip arthroplasties (THAs) with a Spotorno CLS uncemented prosthesis implanted through a Hardinge approach. The mean follow-up was 17.9 years. All consecutive operations were performed by 1 surgeon. Eight patients had been lost to follow-up, and 77 had died for unrelated causes. Overall, 4 stems and 19 cups underwent revision. The cumulative survival rate at 21 years was 79.02% (95% confidence interval [95% CI], 45.98–100.00%) for the acetabular component and 96.71% (95% CI, 60.71–100.00%) for the stem. We can conclude that failure of the Spotorno CLS THA is mainly due to its acetabular component (relative risk 4.5). Survival results for the Spotorno CLS stem exceed the patients' life expectancies in the 60- to 70-year-old population in our area. Loosening with or without fatigue fracture of the component and the learning curve for proper implantation have been the main causes for the expansion cup failure.
Article
Purpose: To search for factors correlated with relapse-free survival following stereotactic reirradiation in patients with recurrent glioma following radiochemotherapy and evaluate tolerance to this treatment. Patients and methods: Initial radiotherapy was given according to the protocol of Stupp and al. Reirradiation was performed using the CyberKnife(®) system. Patients could have had surgical resection initially and at the time of recurrence. We analysed 13 patients treated between July 2010 and September 2014. The median age was 55 years. The doses delivered ranged from 20 to 36Gy, in one to ten fractions. Results: Median survival after stereotactic radiotherapy was 14 months. Survival without relapse was 3.7 months. Factors significantly influencing duration of relapse-free survival were: age (P=0.04), total dose (P=0.02), dose per fraction (P=0.04) and number of fractions (P=0.01). We found no correlation between gross tumour volume, clinical target volume, grade of tumour or prescription isodose and relapse-free survival following radiochemotherapy. Three patients developed radionecrosis. Conclusion: Reirradiation under stereotactic conditions is well tolerated. A dose of more than 30Gy delivered in 5 or more fractions seems to prolong relapse-free survival.
Article
PURPOSE: To evaluate carboplatin, etoposide, and bleomycin (JEB) in children with malignant extracranial germ cell tumors (GCTs). PATIENTS AND METHODS: Malignant GCTs in children aged 0 to 16 years were excised without major morbidity or otherwise biopsied. Stage I testicular and some ovarian GCTs were resected and monitored with alpha-fetoprotein (AFP) (“watch-and-wait” approach). Patients with recurrent stage I disease and all other patients received JEB (etoposide 120 mg/m² on days 1 through 3, carboplatin 600 mg/m² on day 2, and bleomycin 15 mg/m² on day 3). Courses were administered every 3 to 4 weeks until remission, and then two more courses were given. Chemotherapy toxicities were assessed using World Health Organization or Brock grading. RESULTS: Between January 1989 and December 1997, 192 patients were registered. Eight were excluded because either there was no histologic diagnosis (n = 3) or chemotherapy was given off-study (n = 5). The remaining 184 patients had germinoma (n = 20), malignant teratoma (n = 55), embryonal carcinoma (n = 1), yolk sac tumor (n = 107), or choriocarcinoma (n = 1). Forty-seven patients were treated with surgery alone, and 137 patients received JEB. The 5-year survival rate in March 1999 for all 184 patients was 93.2% (95% confidence interval [CI], 87.9% to 96.3%); for the 137 JEB-treated patients, it was 90.9% (95% CI, 83.9% to 95.0%), with an event-free survival rate of 87.8% (95% CI, 81.1% to 92.4%). The median follow-up after JEB treatment was 53 months (range, 0 to 109 months); the median number of courses was five (range, three to eight). Site, stage, and AFP level had prognostic significance. Nonfatal hematologic toxicity was common, but deafness and pulmonary and renal toxicities were rare. One child died of a thoracic tumor and bronchopulmonary dysplasia, and another died of acute myeloid leukemia. CONCLUSION: Conservative surgery, a watch-and-wait approach after complete excision, and JEB for those requiring chemotherapy produced high cure rates and few serious complications.
Article
PURPOSE: To determine whether granulocyte colony-stimulating factor (G-CSF; lenograstim) decreases the incidence of febrile neutropenia after induction courses in treatment of childhood non-Hodgkin’s lymphoma (NHL). PATIENTS AND METHODS: Patients were randomized to receive (G-CSF+) or not receive (G-CSF−) prophylactic G-CSF, 5 μg/kg/d, from day 7 until an absolute neutrophil count ≥ 500/μL was sustained over 48 hours, after two consecutive induction courses of cyclophosphamide 1.5 or 3 g/m ² , vincristine 2 mg/m ² , prednisone 60 mg/m ² /d × 5, doxorubicin 60 mg/m ² , high-dose methotrexate 3 or 8 g/m ² , and intrathecal injections (COPAD[M]) on protocols LMB89, LMT89, and HM91 of the French Society of Pediatric Oncology. RESULTS: One hundred forty-eight patients were assessable, 75 G-CSF+ and 73 G-CSF−. Although duration of neutropenia less than 500/μL was 3 days shorter in G-CSF+ patients (P = 10 ⁻⁴ ), incidence of febrile neutropenia (89% v 93% in the first course, 88% v 88% in the second course), durations of hospitalization and antimicrobial therapy, percentages of infections, mucositis, and transfusions were not significantly different. Although the percentage of G-CSF+ patients commencing the following course on day 21 was significantly higher (84% v 68% after the first and 57% v 38% after the second course; P < .05), the median delay between the two courses was only 1 day less in G-CSF+ patients (median delay after first COPAD(M), 19 v 20 days, P = .01; after second, 21 v 22 days, P = not significant). Remission and survival rates were similar in both arms. CONCLUSION: This study demonstrates that G-CSF did not decrease treatment-related morbidity, nor increase the dose-intensity in children undergoing COPAD(M) induction chemotherapy for NHL.
Article
PURPOSE: Fractionated total-body irradiation (HTBI) is considered to induce less toxicity to normal tissues and probably has the same efficacy as single-dose total-body irradiation (STBI) in patients with acute myeloid leukemia. We decided to determine whether this concept can be applied to a large number of patients with various hematologic malignancies using two dissimilar fractionation schedules. PATIENTS AND METHODS: Between December 1986 and October 1994, 160 patients with various hematologic malignancies were randomized to receive either a 10-Gy dose of STBI or 14.85-Gy dose of HTBI. RESULTS: One hundred forty-seven patients were assessable. The 8-year overall survival rate and cause-specific survival rate in the STBI group was 38% and 63.5%, respectively. Overall survival rate and cause-specific survival rate in the HTBI group was 45% and 77%, respectively. The incidence of interstitial pneumonitis was similar in both groups. However, the incidence of veno-occlusive disease (VOD) of the liver was significantly higher in the STBI group. In the multivariate analysis with overall survival as the end point, the female sex was an independent favorable prognostic factor. On the other hand, when cause-specific survival was considered as the end point, the multivariate analysis demonstrated that sex and TBI were independent prognostic factors. CONCLUSION: The efficacy of HTBI is probably higher than that of STBI. Both regimens induce similar toxicity with the exception of VOD of the liver, the incidence of which is significantly more pronounced in the STBI group.
Chapter
An der Orthopädischen Universitätsklinik Mainz wurden seit 1970 Erfahrungen auf dem Gebiet der Knieendoprothetik gesammelt Je nach Ausdehnung des Kniegelenkschadens wurden anfangs tibiale Hemiarthmplastiken (MclntoshScharnierprothesen (Walldius-, Sheehan-, St-Georg-Prothesen) und später Oberflächenersatzprothesen wie unikondyläre Schlitten (Modular-Knie) bzw. bikondyläre Alloarthroplastiken eingesetzt Letztere wurden seit Mitte der 70er Jahre fortwährend optimiert, sodass ander Orthopädischen Universitätsklinik Mainz zunächst Duocondylar- (1971) und später Duopatellar- (1974) bzw. Total-Condylar I–III (1974–1978) und Insall-Burstein-Prothesen (1981–1982) implantiert wurden. Seit 1987 konnten erste Erfahrungen mit einem modularen Kniesystem, den PFC-Modular-Knieendoprothesen der Firma Johnson & Johnson Orthopaedics gesammelt werden.
Chapter
Lung cancer carries a poor prognosis. If radical surgery remains the treatment of choice, it is feasible only in one out of four patients, the disease being diagnosed in the majority of cases in non-operable patients (general status) or at an advanced stage (stages III, IV). Thus external beam radiotherapy and/or chemotherapy is often the exclusive treatment. The poor clinical outcome is linked to limited thoracic local control (<20% at 2 years) and to distant metastases (>50% 2 years). New therapeutic approaches need to be investigated, exploring systemic, locoregional and also purely local treatments.
Article
Purpose of the study High tibial osteotomy (HTO) is a routine procedure for medial gonarthrosis. Mid-term results are known to be satisfactory, but they deteriorate with longer follow-up. The authors present a long term survival analysis of 109 out of 111 consecutive HTO with a minimal potential follow-up of ten years. Material and methods 111 patients were consecutively operated on for isolated primary varus gonarthrosis between 1977 and 1985: 57 men and 54 women, with a mean age of 53 years (range, 27 to 79 years). X-ray measurements were done on stance, hip-ankle view. Global axial deformation was defined as the angle between mechanical axes of femur and tibia. The respective part of congenital and degenerative tibial deformation was assessed according to Dejour. The angle between femoral and tibial bicondylar lines, representing lateral instability, was added to the tibial degenerative deformation to represent the total degenerative deformation. The goal of correction was a 3 to 7 degree mechanical valgus angulation. At the time of bone healing, 82 patients (74 per cent) had an optimal correction. Two patients were excluded from the follow-up study because of a severe complication (1 bacterial arthritis and 1 tibia non union) which could interfere with the long term result. The 109 remaining patients were followed for a minimal period of 1 year (mean: 8.4 years). GUEPAR pain grading and the occurance of a revision were prospectively analyzed. 57 non reoperated patients could be re-examined at a mean maximal follow-up of 13.5 years (range, 10 to 18 years). Failure was defined as either the occurance of a grade 2 or 3 GUEPAR pain during the whole follow-up, or a clinical or functional Knee Society score < 80 points at final follow-up, or revision. Failure and revision rates were calculated according to Kaplan and Meier. Results 11 patients were reoperated on before final examination (10 per cent): 2 medial unicondylar and 9 total knee prostheses. At final follow-up, the mean clinical and functional scores were respectively 87.0 points (range, 24 to 100 points) and 86.3 (range, 45 to 100 points). The cumulative failure rate was 33 per cent after 10 years and 54 per cent after 15 years; the respective revision rates were 9 per cent and 19 per cent. A pre-operative total degenerative deformation superior to 3 degrees led to a 3.5 fold increased failure rate (p < 0.000,1). A pre-operative medial joint space narrowing over the half of the normal, lateral one led to a 2.2 fold increased failure rate (p = 0.014). An optimal post-operative correction led to a 3.2 fold decreased failure rate (p = 0.000,1). For a given total degenerative deformation, patients with a congenital deformation superior to 5 degrees had a significant lower failure rate (p < 0.000,1). No factor significantly influenced the revision rate. Discussion Ideal patients for HTO, with an expected survival rate of 100 per cent after 13 years, have moderate degenerative changes and a congenital deformation superior to 5 degrees. Patients with advanced degenerative changes and no congenital deformation experienced a 35 per cent failure rate after 10 years. In this population, unicondylar replacement should be considered as a valuable alternative.
Clinical features and mortality due to human immunodeficiency virus type-1 (HIV-1) infection in women are described as part of a prospective 4-year cohort study on perinatal transmission of HIV in Kigali, Rwanda. Two hundred fifteen HIV-seropositive (HIV +) and 216 HIV-seronegative (HIV -) pregnant women were enrolled at delivery between November 1988 and June 1989. Clinical information collected during systematic quarterly examinations was compared. HIV antibody tests were performed at delivery and CD4/CD8 lymphocyte counts at 15 days' postpartum. HIV - women who seroconverted during the follow-up period were excluded from the analysis of the comparison group starting at the date of seroconversion. At enrollment, all HIV + women were asymptomatic for acquired immune deficiency syndrome (AIDS). Incidence of tuberculosis was 2.9 per 100 women-years (WY) after 4 years of follow-up in HIV+ women versus 0.2 per 100 WY among HIV- women (relative risk, 18.2; 95% confidence interval 2.4-137.0). Among HIV + women, the incidence of AIDS (World Health Organization clinical AIDS definition) was 3.5 per 100 WY. The mortality rate was 4.4 per 100 WY among HIV + women versus 0.5 per 100 WY among HIV- women. Clinical AIDS was present in only half of the fatalities. Tuberculosis was a major cause of morbidity and mortality in these HIV + African women. An early diagnosis and an appropriate treatment or prevention of tuberculosis should improve the quality of life of HIV-infected patients in Africa.
Article
We reviewed the outcome of 422 primary cemented Kinemax total knee arthroplasties implanted into 369 patients over a period of five years, from January 1989. The operations were carried out at two NHS district general hospitals and one teaching hospital by 31 surgeons. During the period of review, 49 patients died and ten knees were lost to follow-up (68 knees). The mean Knee Society score improved from 28 before to 89 after surgery, and the mean function score increased from 23 to 79. The range of flexion improved from 92° to 105°. These improvements were maintained throughout the period of study. At the latest review radiolucent lines of 1 mm were seen around 15% of tibial components, 1.4% of patellar components and 9.5% of femoral components. In no case were these changes progressive. Using revision as the endpoint, cumulative survival was 99% after five years and 96.95% after nine years. All revisions were undertaken for deep infection or secondary trauma. Our study has shown that the Kinemax total knee replacement, when carried out with retention of the posterior cruciate ligament by surgeons of varying experience, produces very satisfactory results in the medium term.
Article
High-risk neuroblastoma is characterised by poor long-term survival, especially for very high-risk (VHR) patients (poor response of metastases after induction therapy). We report the results of an intensified high-dose chemotherapy (HDC) strategy to improve the prognosis of VHR patients. This strategy was based on tandem HDC with thiotepa and busulfan-melphalan (Bu-Mel) followed by autologous stem cell transplantation (ASCT). All data were prospectively recorded in the Gustave Roussy Paediatric ASCT database. From April 2004 to August 2011, 26 patients were eligible for tandem HDC. The median age at diagnosis was 4.4 years (1-15.9). All patients had metastatic disease. MYCN was amplified in 5/26 tumours. Despite the cumulative toxicity of alkylating agents, the toxicity of the intensified HDC strategy was manageable. Thiotepa-related toxicity was mainly digestive, whereas sinusoidal obstruction syndrome was the main toxicity observed after Bu-Mel. The 3-year event-free survival of this cohort was 37.3% (21.3-56.7). This strategy will be compared with combined (131)I-mIBG/Bu-Mel in the upcoming SIOPEN VHR Neuroblastoma Protocol.Bone Marrow Transplantation advance online publication, 2 November 2015; doi:10.1038/bmt.2015.264.
Article
Purpose: The role of pelvic elective nodal irradiation (ENI) in the management of prostate cancer is controversial. This study analyzed the role of pelvic radiation therapy (RT) on the outcome in high-risk localized prostate cancer patients included in the Groupe d'Etude des Tumeurs Uro-Genitales (GETUG) 12 trial. Methods and materials: Patients with a nonpretreated high-risk localized prostate cancer and a staging lymphadenectomy were randomly assigned to receive either goserelin every 3 months for 3 years and 4 cycles of docetaxel plus estramustine or goserelin alone. Local therapy was administered 3 months after the start of systemic treatment. Performance of pelvic ENI was left to the treating physician. Only patients treated with primary RT were included in this analysis. The primary endpoint was biochemical progression-free survival (bPFS). Results: A total of 413 patients treated from 2002 to 2006 were included, of whom 358 were treated using primary RT. A total of 208 patients received pelvic RT and 150 prostate-only RT. Prostate-specific antigen (PSA) concentration, Gleason score, or T stage did not differ according to performance of pelvic RT; pN+ patients more frequently received pelvic RT than pN0 patients (P<.0001). Median follow-up was 8.8 years. In multivariate analysis, bPFS was negatively impacted by pN stage (hazard ratio [HR]: 2.52 [95% confidence interval [CI]: 1.78-3.54], P<.0001), Gleason score 8 or higher (HR: 1.41 [95% CI: 1.03-1.93], P=.033) and PSA higher than 20 ng/mL (HR: 1.41 [95% CI: 1.02-1.96], P=.038), and positively impacted by the use of chemotherapy (HR: 0.66 [95% CI: 0.48-0.9], P=.009). There was no association between bPFS and use of pelvic ENI in multivariate analysis (HR: 1.10 [95% CI: 0.78-1.55], P=.60), even when analysis was restricted to pN0 patients (HR: 0.88 [95% CI: 0.59-1.31], P=.53). Pelvic ENI was not associated with increased acute or late patient reported toxicity. Conclusions: This unplanned analysis of a randomized trial failed to demonstrate a benefit of pelvic ENI on bPFS in high-risk localized prostate cancer patients.
Article
To report survival and morbidity of a large homogeneous cohort of patients with a locally advanced esophageal or cardia carcinoma and put in evidence predictive factors of locoregional control and survival. Hundred and two patients were treated at the university hospital of Tours between 1990 and 2010 and received neo-adjuvant chemoradiation therapy with external irradiation (40Gy-44Gy) and two courses of chemotherapy (5-fluoro-uracile and cisplatine). Esophagectomy associated with lymph node dissection was performed about ten weeks after the end of chemoradiation therapy. The median follow-up was 22.4 months [6-185 months]. The overall survival rates at 2 and 5years were 53% and 27%, respectively. The median overall survival was estimated at 27months. The overall 2-year survival between patients "responders" and patients "non-responders" was 67% vs 26%, respectively (P<0.0001). In case of histological response, there was a benefit in terms of overall survival (P<0.0001), locoregional control (P<0.0036) and disease-free survival (P<0.001). Overall survival at 2years was 64% for ypN0 group vs 32% for ypN1 group (P<0.0001). The median survival was estimated at 37months against 15months in the absence of lymph node involvement (P<0.0001). Our results in terms of survival, tolerance and morbidity and mortality were comparable to those in the literature. Complete histological response of lymph node was associated with an improvement of local control, disease-free survival and overall survival. Copyright © 2015. Published by Elsevier SAS.
Article
We present the results of the TGM-95 study for gonadal sex-cord stromal tumors (SCT). Between 1995 and 2005, children (<18 years) with gonadal SCT were prospectively registered. Primary gonadal resection was recommended whenever feasible. Patients with disseminated disease or an incomplete resection received neoadjuvant or adjuvant VIP chemotherapy (etoposide, ifosfamide, cisplatinum). Thirty-eight children with ovarian SCT were registered. Median age was 10.7y. Endocrine symptoms were present in 21 cases. The histological diagnoses were as follows: juvenile (23) and adult (3) granulosa cell tumors, Sertoli-Leydig cell tumors (11), and mixed germ cell SCT (1). An initial oophorectomy ± salpingectomy led to complete resection in 23 patients who did not receive adjuvant treatment; two of them relapsed: one achieved second complete remission whereas the other one died of disease. Fifteen patients had tumor rupture and/or malignant ascites: 11 received chemotherapy and did not relapse, four did not receive chemotherapy and relapsed with a fatal outcome in two cases. With a median follow-up of 5.9y, the 5-y EFS and OS rates were respectively 85% and 94%. Eleven patients had localized testicular tumors (median age 0.83y): juvenile granulosa cell tumors (4), Sertoli or Leydig cell tumors (5) and not otherwise specified SCT (2). Treatment was surgery alone with an inguinal orchiectomy. None have relapsed (median follow-up: 5.4y). Childhood SCT carry favorable prognosis. In ovarian SCT, surgery should be complete and non-mutilating. Adjuvant chemotherapy efficiently prevents recurrences in cases of tumor rupture. In childhood testicular SCT, the prognosis is excellent with an inguinal orchiectomy, prompting the debate on testis-sparing surgery. Pediatr Blood Cancer © 2015 Wiley Periodicals, Inc. © 2015 Wiley Periodicals, Inc.
Article
Full-text available
T HERE ARE three classes of compounds which are known to be effective in the treatment of acute leukemia: adrenal corticosteroids and adreno- corticotrophic hormone (ACTH); purine analogs, such as 6-mercaptopurine (6-MP); and the folic acid antagonists, such as Methotrexate (MTX). There are a number of problems relating to the optimal use of these agents which remain unanswered. This is a report of a study in man of certain of these problems with emphasis on the following: 1. To determine whether the sequence of antimetabohite therapy influences response. In both microbiological and rodent tumcr systems, the development of resistance to one antimetabohito may markedly increase response (collateral sensitivity) to another antimetaholite. 2. To determine whether a patient's response to antimetabohites becomes progressively poorer with increasing disease duration. It is generally accepted that increasing disease duration and prior therapy will adversely affect re- sponse to therapy. While this phenomenon certainhy can be demonstrated in transplanted mouse tumors' the evidence that it occurs in acute leukemia or other human malignant neoplasms and the degree to which it occurs is un- known.
Article
The efficacy of three therapeutic programs for acute leukemia were compared. These programs included (1) Methotrexate (Phase I) followed by 6-mercaptopurine (Phase II); (2) 6-mercaptopurine (Phase I) followed by Methotrexate (Phase II); and (3) Combination Therapy, i.e., 6-mercaptopurine given in combination with Methotrexate. In children with acute lymphocytic leukemia the remission rate was 59 per cent for combination therapy, 47 per cent for 6-mercaptopurine, and 29 per cent for Methotrexate. The better remission rate for combination therapy is consistent with that predicted if it is assumed that 6-mercaptopurine and Methotrexate act independently. The median duration of complete remissions for the three treatments was not different (4 to 5 months). However, long lasting remissions were more frequent in patients receiving combination therapy. The median survival from the onset of therapy to death was 9 months. There were no differences between the three treatment programs as regards survival. In adults the remission rate was 15 per cent for combination therapy, 21 per cent for 6-mercaptopurine and 7 per cent for Methotrexate. As regards survival in adults, early deaths were more common in patients who received MTX as initial therapy, whereas after 5 months survival was somewhat better in those patients receiving combination therapy. In both children and adults there was no evidence that prior treatment with one of the antimetabolites altered response to the other antimetabolite. This result differs from those in animal models, and its effect on our concept of the mechanism of resistance is discussed. Responsiveness to the second course of antimetabolite therapy (Phase II) was as good as that to the first course of treatment. This was true for the remission rate, remission duration, and even for survival when appropriate corrections were made. Thus, responsiveness to drug therapy is maintained as the disease progresses temporally. It may be concluded therefore that new agents can be effectively studied in patients with "late" disease. Responsiveness to Phase II therapy was independent of responsiveness to Phase I. The most common and severe toxic manifestations related to the bone marrow and gastrointestinal tract. There were no major differences quantitatively in the toxicity for the three treatment programs in children in spite of the fact that the drugs were given in full dosage in the combination program. Oral ulcers and a generalized erythematous rash occurred significantly more frequently in patients receiving Methotrexate. Jaundice was significantly more frequent in adults and in patients receiving 6-MP.
Article
We have illustrated the life table method for computing survival rates with 5-year survival data for cancer patients, emphasizing the advantage gained by including survival information on cases which entered the series too late to have had the opportunity to survive a full 5 years. The advantage is measured in terms of reduction in standard error of the survival rate. For the five series of patients in this paper, the reduction in standard error ranged from one-third to two-thirds.
Appendix 1. The " Error of Sampling
  • Greenwood
Greenwood M: Reports on Public Health and Medical Subjects, No. 33, Appendix 1. The " Error of Sampling " of the Survivorship Tables, H.M. Stationery Office, London, 1926
The “Error of Sampling” of the Survivorship Tables
  • Greenwood