Clinical data missing

There may be times when a DATA step with arrays is a better means to transpose data. This is true when the data to be transposed have more than one record per BY group variable or when there is a need to have the resulting data set include data that are not in the source data set. In clinical trials missing data is a very common issue. Let s look at a derivation of the previous systolic blood pressure transposition problem where visit 2 is always missing. 

In this case, about half of the data desired were available electronically, most of which were related to health as heavily weighted toward economic information. To gather the remaining desired data the investigators needed to collect prospectively humanistic as well as some additional clinical variables (Hirsch and Van Den Eeden, 1997). It is quite typical that clinical data available electronically are often not complete and therefore not very useful, and that humanistic data are missing completely from the databases held by Health Maintenance Organizations. 

Dearth of patient-centered outcomes measures in traditional drug development. Physicians are usually relied upon for clinical data. Data from the patient are sometimes perceived to be soft . Patient perspectives may also be missed when the traditional clinical focus may be disease- or organ-dominated. 

Iodine supplementation probably begins late in many pregnant women (at the first prenatal care visit), missing the critical period of fetal brain development. However, there is no clinical data so far that supports the effect of iodine supplementation on neural development in pregnant women from mild iodine-deficient areas in both the short- and long-term. These data indicate that continued efforts for USI remain a public health priority. 

Prospective sources include encounter data, which may or may not be contained in EHRs patient data input and randomized, prospective clinical trials. Advantages of prospective sources to inform interactive software include the ability to control and monitor the circumstances of data collection reduction (as a result of randomization) of sources of bias potential minimization of missing data potential to modify design of data collection ability to verify data accuracy and ability to validate and further test assumptions and modify existing programs. 

Missing Information. Both premarketing and postmarketing collections of data are perpetually plagued by missing information. In premarketing and postmarketing clinical trials, patients can be lost to follow-up because of ... 

Although 500,000 individuals were enrolled in clinical trials that were submitted to the FDA during 1990-1995 [10], the lack of a repository of clinical trial data, standardized data, and interoperable systems precludes us from efficiently tapping and reanalyzing these data. This missed opportunity underscores the need for standardization and interoperable systems, as discussed above (see Section 27.4.1 on data standards and interoperable systems). 

Notice the missing column for visit 2. This is exactly what you would expect PROC TRANSPOSE to give you. PROC TRANSPOSE transposed the data that were present and could not be expected to know about visits that are not represented in the data. However, often in clinical trials reporting you want to report on all visits, treatments, or other expected parameters whether they are represented in the actual data or not. In this case, a DATA step with arrays is a better choice to transform the data. Here is an example of the previous transposition that includes all visits 1-5, regardless of which visits are included in the underlying data. 

Using SQL, you can create very sophisticated database joins that were not easily done before SQL became part of Base SAS. Keep in mind that these data set joins can get very complex. For example, the previous scenario gets quite a bit more complicated with the introduction of missing start and stop dates, which are a common occurrence in clinical trial data. 

Timing of any interim analysis Reason for sample size Power of clinical study Level of significance Criteria for termination of study Procedure for missing and imused data Reporting procedures for deviations from statistical plan... 

One issue that is important to determine is how missing data will be handled in an ITT analysis. ITT was largely developed in clinical trials in which the major endpoints were events, mortality, infarctions, etc. In such studies it is possible to follow-up on patients who withdraw from treatment and determine whether the event has occurred or not and indeed in such trials every effort should be made to do this. For other types of data, for example, a pain scale other approaches are necessary. 

Molenberghs, G. and Kenward, M., 2007, Missing data in clinical studies, John Wiley Sons. 

The purpose of our study was to produce only a mild zlnc-deflclent state In human volunteers. Inasmuch as severe deficiency of zinc may be llfe-threatenlng, as seen In acrodermatitis enteropathlca. Furthermore, It Is the marginal deficiency of zinc that appears to be prevalent and likely to be missed clinically. Our data show that we were successful In producing a mild zlnc-deflclent state In human volunteers by dietary means. Zinc concentration of plasma, erythrocytes, leukocytes, and urine decreased when the dally Intake of zinc was restricted to... 

---