Analysis of Multivariate Survival Data

Hougaard, Philip

doi:10.1007/978-1-4612-1304-8

Cited by 1,166 publications

(1,082 citation statements)

References 0 publications

Supporting

Mentioning

1,042

Contrasting

Unclassified

Order By: Relevance

“…For this purpose, episodes were analyzed as recurrent events using a gamma frailty model. 16 All tests, except for the analysis of the primary end point, were 2 sided. P values <0.05 were considered statistically significant.…”

Section: Discussionmentioning

confidence: 99%

Comparison of insulin detemir and insulin glargine in a basal—bolus regimen, with insulin aspart as the mealtime insulin, in patients with type 1 diabetes: A 52-week, multinational, randomized, open-label, parallel-group, Treat-to-Target noninferiority trial

Heller

Koenen

Bode

2009

Clinical Therapeutics

View full text Add to dashboard Cite

Objective: The primary study objective was to determine whether insulin detemir (detemir) was noninferior to insulin glargine (glargine) as the basal insulin in a basalbolus regimen, with insulin aspart as the mealtime insulin, in terms of glycemic control at the end of 52 weeks in patients with type 1 diabetes mellitus (T1DM). Methods:This multinational, open-label, parallel-group, treat-to-target, noninferiority trial enrolled patients aged ≥18 years who had had T1DM for at least 12 months, had been taking a basal-bolus insulin regimen for at least 3 months, and had a glycosylated hemoglobin (HbA 1c ) value ≤11.0%. Patients were randomized in a 2:1 ratio to receive either detemir given once or twice daily or glargine given once daily for 52 weeks. The basal insulin was initially administered once daily (in the evening) in both groups; if patients in the detemir group were achieving the PG target before breakfast but not before dinner, they were switched to twice-daily administration. Each patient attended 13 study visits and received 16 scheduled telephone calls from the trial site.The primary efficacy end point was glycemic control (HbA 1c ) after 52 weeks of treatment. Secondary end points included the number of patients achieving an HbA 1c value ≤7.0%, with or without a major hypoglycemic episode in the last month of treatment; fasting plasma glucose (FPG); within-patient variation in self-monitored plasma glucose (SMPG) before breakfast and dinner; and 10-point SMPG profiles. The noninferiority margin was 0.4%, consistent with US Food and Drug Administration guidelines. significantly between the detemir and glargine groups (7.57% and 7.56%, respectively; mean difference, 0.01%; 95% CI, -0.13 to 0.16), consistent with the noninferiority of detemir to glargine. The corresponding estimated changes in HbA 1c were -0.53% and -0.54%. In the 90 patients who completed the trial on once-daily and the 173 patients who completed the trial on twice-daily detemir, the estimated changes in HbA 1c were -0.45% and -0.56%, respectively. After 52 weeks, there were no significant differences in the proportion of those receiving detemir and glargine who achieved an HbA 1c value ≤7.0% without major hypoglycemia (31.9% and 28.9%, respectively). In addition, there were no significant differences in estimated mean FPG (8.58 and 8.81 mmol/L; mean difference, -0.23; 95% CI,-1.04 to 0.58) or in basal insulin doses. The basal insulin dose was numerically higher in patients receiving detemir twice rather than once daily (0.47 vs 0.33 U/kg, respectively). The relative risks for total and nocturnal hypoglycemia with detemir versus glargine were 0.94 and 1.12, respectively (both, P = NS). Six patients (2.0%) in the detemir group and 4 (2.8%) in the glargine group withdrew due to adverse events. Results Conclusion

show abstract

Section: Discussionmentioning

confidence: 99%

Comparison of insulin detemir and insulin glargine in a basal—bolus regimen, with insulin aspart as the mealtime insulin, in patients with type 1 diabetes: A 52-week, multinational, randomized, open-label, parallel-group, Treat-to-Target noninferiority trial

Heller

Koenen

Bode

2009

Clinical Therapeutics

View full text Add to dashboard Cite

show abstract

“…Standard descriptive and non-parametric comparative statistics were used to characterise and compare groups. The number of severe hypoglycaemic episodes and the number of severe hypoglycaemic episodes with coma were analysed by a frailty model for recurrent events (an extension of the log-linear Poisson model including a gamma-distributed variation between patients) that takes into account the highly skewed distribution of events of severe hypoglycaemia [18]. The effect of an explanatory factor is reported as relative risk with 95% confidence limits.…”

Section: Methodsmentioning

confidence: 99%

Prediction of Severe Hypoglycaemia by Angiotensin-Converting Enzyme Activity and Genotype in Type 1 Diabetes

et al. 2003

View full text Add to dashboard Cite

Aims/hypothesis. We have previously shown a strong relationship between high angiotensin-converting enzyme (ACE) activity, presence of the deletion (D) allele of the ACE gene and recall of severe hypoglycaemic events in patients with Type 1 diabetes. This study was carried out to assess this relationship prospectively. Methods. We followed 171 adult outpatients with Type 1 diabetes in a one-year observational study with the recording of severe hypoglycaemia. Participants were characterised by serum ACE activity and ACE genotype and not treated with ACE inhibitors or angiotensin II receptor antagonists.Results. There was a positive relationship between serum ACE activity and rate of severe hypoglycaemia with a 2.7 times higher rate in the fourth quartile of ACE activity compared to the first quartile (p=0.0007). A similar relationship was observed for the subset of episodes with coma (2.9 times higher rate in fourth quartile compared to first quartile; p=0.048). The impact of serum ACE activity was most pronounced in C-peptide negative subjects (4.2 times higher rate in fourth quartile compared to first quartile; p=0.003), and in this subgroup carriers of the D allele of the ACE gene had higher rates of severe hypoglycaemia compared to the group homozygous for the insertion (I) allele. In a multiple regression analysis high serum ACE activity and impaired awareness of hypoglycaemia were identified as the only significant predictors of severe hypoglycemia. To avoid progression of these frequent episodes into severe hypoglycaemia patients rely on maintenance of performance capacity [3] until hypoglycaemic symptoms are recognised and self-treated or until the next pre-scheduled meal is ingested.Risk of severe hypoglycaemia is a major limiting factor for the achievement of good metabolic control [4] rendering prediction of individual risk of severe hypoglycaemia increasingly important. Loss of hypoglycaemic warning symptoms and blunted hormonal counterregulatory response are the most important known predictors of severe hypoglycaemia [4,5]. However, even patients at high risk according to these paFollowing the Diabetes Control and Complications Trial [1] a growing demand for strict glycaemic control has emerged to avoid late diabetic complications in Type 1 diabetes. As a consequence the majority of patients with Type 1 diabetes experience mild hypoglycaemia on a regular basis and recent studies applying continuous glucose monitoring indicate that pa-

show abstract

“…This is the approach I now recommend for analysing survival data for AB/BA cross-over trials. However, such methods do not readily generalize to more complex designs and it may be worth studying Philip Hougaard's magnificent book on multivariate survival analysis [31] for alternatives. …”

Section: Time-to-event Analysismentioning

confidence: 99%

Cross‐over trials in Statistics in Medicine: the first ‘25’ years

Senn

2006

Statistics in Medicine

View full text Add to dashboard Cite

SUMMARYPapers on cross-over trials that have appeared in the first 25 years of Statistics in Medicine are reviewed. Papers on bioequivalence are also considered. After a brief statistical summary, individual papers are discussed under seven headings: 1. The two-stage analysis of AB/BA trials, 2. Baselines, 3. Binary and categorical data, 4. Survival data, 5. Modelling carry-over, 6. Bioequivalence and 7. Components of variation. Finally, a brief assessment of the importance in this field of Statistics in Medicine is given.

show abstract

Analysis of Multivariate Survival Data

Cited by 1,166 publications

References 0 publications

Comparison of insulin detemir and insulin glargine in a basal—bolus regimen, with insulin aspart as the mealtime insulin, in patients with type 1 diabetes: A 52-week, multinational, randomized, open-label, parallel-group, Treat-to-Target noninferiority trial

Comparison of insulin detemir and insulin glargine in a basal—bolus regimen, with insulin aspart as the mealtime insulin, in patients with type 1 diabetes: A 52-week, multinational, randomized, open-label, parallel-group, Treat-to-Target noninferiority trial

Prediction of Severe Hypoglycaemia by Angiotensin-Converting Enzyme Activity and Genotype in Type 1 Diabetes

Cross‐over trials in Statistics in Medicine: the first ‘25’ years

Contact Info

Product

Resources

About