Type I error rates from likelihood‐based repeated measures analyses of incomplete longitudinal data

The last observation carried forward (LOCF) approach is commonly utilized to handle missing values in the primary analysis of clinical trials. However, recent evidence suggests that likelihood‐based analyses developed under the missing at random (MAR) framework are sensible alternatives. The objective of this study was to assess the Type I error rates from a likelihood‐based MAR approach – mixed‐model repeated measures (MMRM) – compared with LOCF when estimating treatment contrasts for mean change from baseline to endpoint (Δ). Data emulating neuropsychiatric clinical trials were simulated in a 4 × 4 factorial arrangement of scenarios, using four patterns of mean changes over time and four strategies for deleting data to generate subject dropout via an MAR mechanism. In data with no dropout, estimates of Δ and SE_Δ from MMRM and LOCF were identical. In data with dropout, the Type I error rates (averaged across all scenarios) for MMRM and LOCF were 5.49% and 16.76%, respectively. In 11 of the 16 scenarios, the Type I error rate from MMRM was at least 1.00% closer to the expected rate of 5.00% than the corresponding rate from LOCF. In no scenario did LOCF yield a Type I error rate that was at least 1.00% closer to the expected rate than the corresponding rate from MMRM. The average estimate of SE_Δ from MMRM was greater in data with dropout than in complete data, whereas the average estimate of SE_Δ from LOCF was smaller in data with dropout than in complete data, suggesting that standard errors from MMRM better reflected the uncertainty in the data. The results from this investigation support those from previous studies, which found that MMRM provided reasonable control of Type I error even in the presence of MNAR missingness. No universally best approach to analysis of longitudinal data exists. However, likelihood‐based MAR approaches have been shown to perform well in a variety of situations and are a sensible alternative to the LOCF approach. MNAR methods can be used within a sensitivity analysis framework to test the potential presence and impact of MNAR data, thereby assessing robustness of results from an MAR method. Copyright © 2004 John Wiley & Sons, Ltd.     

Large scale communication networks—Design and analysis

The paper discusses the problem of design and analysis of large scale communication systems. An iterative process, composed of a minimum cost network design algorithm and a network performance algorithm, is presented for the solution of these problems. Computational considerations, using these algorithms, are discussed.     

Assessing the effectiveness of strategic planning

This paper describes a method for assessing the effectiveness of strategic planning systems in terms of the results actually observed subsequent to planning activity. The paper explains in particular how, in making the assessment, one might cope with the fact of environmental change subsequent to the setting of planned goals. This ends oriented approach to measuring effectiveness is shown to be complementary to process oriented approaches and is illustrated by an example from the coal industry.     

Improving homeless persons' utilisation of primary care: lessons to be learned from an outreach programme in The Netherlands

Elissen AMJ, Van Raak AJA, Derckx EWCC, Vrijhoef HJM. Improving homeless persons' utilisation of primary care: lessons to be learned from an outreach programme in The Netherlands Faced with rising homelessness, countries around the world are in need of innovative approaches to caring for those without shelter, who, more often than not, suffer from severe health problems. We conducted a case study of an innovative Dutch Primary Care for the Homeless (PCH) programme to gain insight into clients' demographic characteristics, health problems and service use, and to develop an explanation for its success in increasing the latter. Our analyses are based on a combination of quantitative and qualitative data. The results of the study suggest that the success of the PCH programme can be explained by the providers' pragmatism and will to adapt their mode of care provision to the behavioural patterns and needs of their homeless clients.     

A case study comparing a randomized withdrawal trial and a double-blind long-term trial for assessing the long-term efficacy of an antidepressant

Assessing long-term efficacy in psychiatric drugs involves a number of complex questions, and the priaority of these questions is different for different disorders and for different stakeholders. Therefore, it is essential that we not adopt a one-method-fits-all approach, but rather adapt the specific details of the designs and analysis of data from long-term trials to individual disease states. Randomized withdrawal (RW) designs, even though addressing a specific question of particular interest, face some difficult methodological and ethical challenges. A less common alternative design, termed the double-blind long-term efficacy (DBLE) design, is logistically similar to traditional responder extension designs. However, use of an analytic approach that includes all randomized patients rather than only the selected subset that continued beyond acute treatment avoids the major criticism of the extender design. The present paper illustrates the attributes of the RW and DBLE designs by analyzing data from trials adopting these designs. The RW and DBLE designs address different questions, and are thus not directly comparable. Potential benefits of the DBLE design include: (1) the parsimonious use of patients and the resultant reduced exposure to placebo as each patient can contribute to multiple developmental objectives; (2) the results are generalizable to actual clinical practice as the design matches treatment guidelines; and, (3) results of safety assessments are meaningful as they involve all randomized patients. Our case study suggests that the DBLE design can provide definitive answers to important questions and may be a useful design for assessing long-term treatment effects.