“In Hollywood now when people die they don’t say, “Did he leave a will?” but “Did he leave a diary?”” — Liza Minelli

The importance of collecting patient reported outcome data in biopharma research is increasing rapidly. The recent geometric explosion of technological innovation has brought about the introduction and use of electronic diaries to fulfill this purpose. We have reached a point when it is no longer cost-prohibitive or technologically infeasible to collect almost any type of data via electronic means, but this capability has outpaced our operational methodologies and capabilities for the successful large-scale implementations now required. This is one important reason why electronic diary adoption is so much slower than EDC at the moment.

The broad term used for the collection of patient-reported (as distinct from investigator-recorded) data is ePRO (Electronic Patient Reported Outcomes), which encompasses a wide range of so-called “instruments” (series of questions, not hardware) and instruments (hardware) administered in a variety of ways (completed in real-time, or otherwise; completed by patient directly, or otherwise; etc.). In this column, however, I want to focus on the most challenging (but often most important) format of ePRO – the electronic diary.

While use of the common telephone is a common resource for ePRO when data collection requirements are simpler, a full-fledged patient diary requires technology in the handheld and tablet format, where it seems that every few months a new gadget emerges that can do something which simplifies one’s life and makes it that much more complicated at the same time. The data collection options are no longer limited to answering a question. We are now able to visualize the question textually and diagrammatically before providing an answer. We can easily record the exact time and date the answer was provided, limit when and how often the answer can be given, send reminders to prompt for that answer and if we want to get really personal, record the exact location of where that answer was given and have a photograph taken of the person giving the answer. These types of mainstream technological enhancements may pave the way for potential ad hoc analyses of how location could affect an answer (temperature, climate, urban, rural), or provide additional biometric markers in relation to the emotion of the person at the time they were giving an answer. I will not attempt to address the scientific validity of any of these measures, rather the point is to illustrate that enhanced capability and operational complication are directly related, and the latter is often overlooked in favor of the former.

It’s So Easy

What could be so difficult about answering a question on a device which asks you to push a button or select an option with a stylus? This is a common view, and a common misperception of technology providers and sponsors using ePRO. We are so used to dealing directly with sites and personnel who are experienced in the conduct of clinical research that we forget our nice new gadget is now in the hands of the general public, who are by definition not in the research data management mainstream. Not only have we failed to adequately address this fact, but rather have exacerbated the problem by not providing adequate training on how to use these devices for our trials. The training, as with other technologies (EDC), is often considered to be of ancillary importance. We hope that in one, three-hour, session at an investigator meeting, not only will sites become experts on the ePRO device and survey use, but they will also be able to convey this knowledge in a simplified manner to the inexperienced end user. As expedient as this is to sponsor and vendor, it is unfortunately not sufficient to the public user upon whom the data (and study results) now depend.

To further complicate matters, the tendency has been to take full advantage of the capability of the technology to collect ultimately extraneous data. Once again we forget the end user and direct our attention to the almost limitless potential to collect multiple points of data for theoretical ad hoc analyses, often losing site of the primary datapoints we are interested in. It is a classic case of allowing the details and options to confound and overwhelm the intended purpose. I am reminded of the time when a relative of mine from South Africa first walked into a supermarket in the USA in search of a loaf of bread. She was so used to having only two choices — “wheat or white” — that she was overwhelmed by the number of choices available to her and walked out without buying. The second time she went, she returned with fifteen loaves of all different breads — one was eaten and the rest spoiled. And so too it is easy to lose focus when presented with multiple possibilities and end up with lots of extraneous, unutilized data that took a large amount of effort to collect.

Where in the World is Carmen’s Diary?

Globalization is another area which has been recognized as a significant challenge in the implementation of ePRO. Our diary writers now live all over the world, speak many different languages and is recording data every hour of every day of the year. There are two primary challenges in getting patients to use the diary correctly: 1) actually getting the device to them in the first place and, 2) providing them support once they begin using the device.

The provisioning challenge stems directly from the fact that despite all of the innovation we are still using a local piece of hardware that has to be shipped, inventoried and repaired or replaced if damaged. Using ePRO globally will require that sufficient lead times are considered to address issues such as, shipping, customs, and local couriers, not to mention algorithmic calculations to allow for adequate inventory management related to expected screen failures and completed patients. This inventory balance is especially difficult considering the cost replacement (device cost, image loading costs, QA) and the importance placed on missing pieces of data that are irretrievable and thus un-analyzable. It is evident that one cannot record data in the device if they do not have access to it and the technology loses its primary benefit of real time recording, if the contingency is to back enter data once a new device becomes available.

One may ask, “gee, sounds like EDC in the ‘90’s; the Internet took care of that, didn’t it?”. For obvious reasons, the Internet does not quite yet supersede dedicated hardware for ePRO. If protocols call for true, frequent, real-time recording of data, then reliable Internet access – even through enabled smartphones – is not dependable in most areas of the world (try running errands in your hometown all day and think about hourly recording of your symptoms through your smartphone being dependent on a WiFi connection).

I Just Called to Say…I Don’t Understand You.

Another crucial aspect of globalizing ePRO is providing ongoing support to subjects for any questions or technical difficulties that may arise. The difficulty here cannot be exaggerated. It is a very different proposition to be supporting the average trial subject in the use of technology than it is to support clinical research professionals (monitors, etc., as we are used to). There are few if any large helpdesk providers that are familiar enough with this technology to be effective, and supporting an endless number of languages within a manageable cost structure is next to impossible. Vendors have tried to solve this problem by locating their helpdesks in exotic locations claiming multi-language support, when in fact the focus is more on cost-savings due to lower hourly rates. The result is a frustrated subject who spends the majority of the phone call trying to find a person that s/he can understand. This frustration typically leads to lower compliance and a real hesitancy on the part of the sponsor and to use the technology again.

Lead Me to the Landfill

Technological innovation, for all its positives, has left us with an important underlying problem: technology outpaces clinical trial timelines. This puts sponsors in the position of having to continually re-purchase new hardware as the device they used for their previous trial is no longer supported or obsolete. Many argue that this is the cost of doing business and that the capital expense is “the pill you swallow” for the benefit of real time and accurate data. However, in the current economic environment, vendors should be thinking about options that are not hardware-dependent and that can be leveraged across trials. Not only would this go a long way in helping solve part of the provisioning problem, but would also allow some recognizable cost-savings that can be realized across large programs, which currently is a barrier to sponsor acceptance. There is no question that for the most part, for the new generation of clinical research subjects, the technology learning curve will decline. Web access, and its inherent detachment from any one piece of hardware, will undoubtedly help the current hardware dependencies but the solution is not immediately at hand, and sponsors need to be using ePRO now. So much work remains to be done by vendors and their customers alike.

There is little question of the value of collecting patient reported outcomes electronically, and it is clear that the industry in general is moving towards this methodology in the same way they have been doing with EDC for the past decade. Our excitement and acceptance of the innovations of today and tomorrow should not distract us from the importance of operational changes and procedures that play a significant role in realizing the benefit of this innovation. If we can achieve operational excellence in ePRO, we have a chance to overcome temporary barriers to ePRO acceptance.