Data Dilettantes

The success or failure of a clinical trial turns on the integrity of the data, i.e. whether it is complete, accurate, and contemporaneously-recorded. The integrity of the data depends on the diligence of those collecting and recording it. And that is where many clinical trials run into trouble.

Most investigators and other site personnel participating in clinical trials start out with the best of intentions. However, good intentions alone will not result in high quality data. They must be supplemented with comprehensive protocol training and frequent follow-ups to prevent the normal daily demands of a busy clinic from distracting the investigator and other site personnel from their important study responsibilities .

And, for many trials, that’s only half the battle! Study subject owners are often charged with administering the test drug, recording dosing times and documenting their observations. Owners must clearly understand the importance of strict adherence to the dosing regimen and complete, contemporaneous records of their observations. A cursory description of the study during the consent discussion does not properly prepare the owner to carry out their critical responsibilities.

When investigators, technicians, and subject owners are fully trained, the results are wonderful. When they are not, the data suffers. Let me give you some examples …

In a study to evaluate the efficacy of a novel compound, the investigator instructed the owner to administer the test article once per day, at approximately the same time . After the owner repeated the instructions to the investigator and also signed the ‘Ow ner Instruction Sheet; the investigator felt confident the owner would be compliant.

What the investigator did not know was that the owner would be out of town for three days and would delegate test article administration to her adult son, who would be taking care of her dog while she was away. As you might have guessed, the owner’s son was himself quite busy. So, although he came to the house to feed and walk the dog several times each day, he forgot to administer the test article.

When the owner returned to the clinic for the second study visit, the investigator reviewed her owner diary and discovered the missing data. Since the protocol limited missed doses to two, this case was removed from the study.

In a different study, the owner devised a method to make dosing more ‘efficient’. Instead of administering the twice-daily dose, as specified in the protocol and as instructed by the investigator, he gave both doses together each morning and duly recorded that fact in his diary.

Since both doses were recorded, one might argue that this is not a case of missing data. However, when looking at the owner diary, the second daily dose had not been documented at the proper time and, therefore, was missing. Either way, the case was removed from the study.

In a third study, the scale used to weigh study subjects was not functioning correctly. The technician realized this when she weighed a small breed dog and the scale showed 63 pounds .

Rather than deviate from the required study visit window, the technician recorded the weight as ‘not available; wrote an explanatory note for the study file, and included the last weight recorded in the subject’s medical record as an approximate weight. Nonetheless, the body weight for this study subject at this visit was ‘missing’.

According to a 2012 article in the New England Journal of Medicine, “Missing data are defined as values that are not available and that would be meaningful for analysis if they were observed”.

Missing data can have serious implications for clinical trials. Sponsors must evaluate each instance and decide whether the affected case should be included in the database or discarded. Incomplete cases that are included in the database may raise concerns during the final review process that could result in regulatory site audits and/or delayed approval of the test drug. This risk is very real because there are significant financial consequences to such delays.

While it is unrealistic to expect that all data will be properly recorded, there are ways to minimize data loss and improve the overall quality of a study’s database.

First, ensure that all equipment required by the study protocol is functioning properly at each investigative site on every day of the in-life phase of the study. In the event that a malfunction is discovered, use a different, properly-calibrated piece of equipment or, if alternate equipment is not available, reschedule associated subject study visits until the equipment has been fixed.

Second, contact study subject owners between study visits to discuss their experience in administering the test drug, review their diary entries, and address any problems. This is especially important at the beginning of each subject’s participation, to catch owner mistakes before they do enough damage to disqualify a case.

Third, and most important, know thy client! Owners who have not followed prior treatment regimens may not be the best candidates for studies requiring independent administration of the test drug and/or regular documentation of their observations. Owners who are relatively new to the clinic should be vetted carefully. Of course, owners who have complied with previously-prescribed treatments or who are particularly attentive to their pet’s health usually make good study participants.

I say ‘usually’ because all owners are human with other demands on their minds. So, regardless of the investigator ‘s expectation of future owner performance, all owners should receive comprehensive, detailed instruction in their responsibilities under the protocol and be regularly monitored until their performance against the study’s requirements can be ascertained. It also is beneficial to review those responsibilities at each study visit, just as a refresher.

Lastly, it is important to impress upon each owner that, by signing the consent form, they are agreeing not only to their pet’s participation in the study but also to execute their responsibilities in a diligent manner . They may not delegate those responsibilities to anyone else – not a spouse, parent, child, or friend – without notifying the investigator and without that person also being trained in the owner’s responsibilities under the protocol.

Study sponsors have spent years of research effort and millions of dollars to ready a test drug for a clinical trial. They have a lot riding on the outcome of that trial and, consequently, have a right to expect protocol compliance from all participants . All of us should work hard to give them the complete and accurate data they deserve.