What should doctors do with patient-generated health data?

Tuesday, August 23, 2022
Data
News
Patient-Generated Health Data (PGHD) might suggest the correct diagnosis. However, it may also lead to errors if, for example, the patient takes the measurements incorrectly or when the device itself is imprecise. Thus, it's not easy for the doctors to decide what to do with them.

PDFs and print-outs with measurements, trends, and diagrams

New health-monitoring tools – mobile health and wellness applications, smartwatches, and wristbands – allow patients to collect and submit data to health practitioners. On the one hand, it is a positive phenomenon – there is hardly anything more valuable than an engaged patient who deeply cares about their well-being. Diseases such as hypertension or migraine are often diagnosed precisely through the patient's observations and notes. On the other hand, the amount of data can be overwhelming – it is often a dozen pages of printed-out fitness application data, graphs with patients' notes and physical activity trends. Patients also bring heart rate and sleep quality statistics from several months prior, thinking they should be carefully studied. The doctor, following their experience and knowledge, can decide whether the data have any clinical importance. If they do not – they should be ignored because the practitioner has no way of measuring the device's accuracy. However, the patient – who had worked hard to collect the data – is convinced the information is valuable. After all, they had invested a lot of energy (and money) into monitoring their health. As a result, the relationship with the patient deteriorates, and trust is lost. Patients feel ignored, and practitioners – overwhelmed with irrelevant data.

PGHD – helpful or disruptive?

It all depends on the context. However, the conclusions of the available scientific analyses are optimistic. An example would be a study published in 2020 in Health Informatics Journal How patient-generated health data and patient-reported outcomes affect patient-clinician relationships: A systematic review. In the eight publications analyzed, patients and clinicians perceived PGHD as a tool that helped to improve communication. In another five publications, an improvement in relations was described, and two studies concluded that PGHD allows doctors to provide more empathetic care. This element turned out to be very important in the patient-practitioner relationship, as indicated in the study. Aside from that, doctors modified their treatment plans several times after studying PGHD. However, there were also a lot of negative opinions, mainly relating to the lack of time to analyze the data provided by the patient, its non-uniformity, or the lack of guidelines. The lack of integration between PGHD, clinical workflows and Electronic Health Records (EHR), which results in difficulties in qualifying such information, was also raised. Interestingly, there were also situations in which doctors had pointed out the deterioration of communication as they had to focus on reading stacks of papers instead of conducting a medical interview.
The proper procedure can help in a quick evaluation of the PGHD
According to a study published in 2019, Patient Generated Health Data Use in Clinical Practice: A Systematic Review, 46% of Americans were "digitally active" in 2016, meaning they used three or more digital health devices. According to EUROSTAT, almost one-fifth (19%) of EU citizens aged between 16-74 had a smartwatch, fitness band, or other device monitoring their physical activity in 2020. And their sales are also growing because patients want to measure and manage their health. No doubt, PGHD will gain importance since they catch essential, but today ignored data regarding behavior, habits and health parameters captured in home settings. Some studies suggest that PGHD may lead to better clinical decision-making and empowering patients. There is a lot of information coming from the patient:
  • Personal: health goals, values, worries, and hopes
  • Medical history: diseases in the family and hereditary diseases
  • Medication: drugs taken by the patient and not found in the records
  • Biometric data (nowadays measurable at home): blood pressure, weight, body temperature, oxygen saturation, glucose level, heart rate
  • Behavioral data: physical activity, burned calories, sleep quality, other routine activities that impact health (e.g., diet)
  • Environmental data: room temperature, noise, air humidity, place of residence, air quality (pollution)
  • Social interactions: time spent online and real-life interactions
  • Patient's observations regarding their treatment (the so-called patient-generated results)
Most of these are not included in the patient's health records.

Who has time for this?

Healthcare providers worry that the additional responsibility of analyzing PGHD is beyond their capabilities, especially considering the limited time per appointment. In a simulation test conducted to study the impact of PGHD on the healthcare system, the researchers identified such indirect consequences as, for example, increased labor costs due to the time it takes to analyze the data. An indirect result was a disturbed work schedule. In addition, practitioners and nurses declared they were not competent enough to analyze PGHD. The authors of the study pointed out that integrating PGHD in clinical practice will leave less time for other elements of the patient's visit. Healthcare providers are also unsure if the data provided by their patients are useful and of high quality to assist in decision-making. There is another element: responsibility. What if the doctor ignores seemingly insignificant data that actually pointed toward a severe disease? Should the doctor be accused of a medical error? An equally important question is how – if at all – to integrate PGHD into the Electronic Health Records (EHR). And how to effectively analyze the data to support clinical decision-making. The problem with PGHD would disappear if the data were always provided electronically. Artificial intelligence systems could then analyze big data sets in just a few seconds to check if there are any alarming trends.

What can a doctor do today?

Experts recommend every medical facility develop a set of standards that doctors and nurses can follow with a large margin for their own assessment. Of course, it always depends on the relationship between the patient and the doctor, if the patient is new or chronic, etc. The procedures should encompass such steps of decision-making as:
  • Is the data clinically relevant?
  • If yes, how was it collected? Is it reliable?
  • Did the patient take the measurements correctly?
  • If the data is not clinically significant, can it be used in other contexts, e.g., prevention?
  • What information should be included in the EHR? If so, where? Who should be responsible for the data (doctor, nurse)?
  • How to communicate with the patient while providing the self-collected data?
  • In what context is the data provided? Are there other symptoms, doubts, or observations relevant to the information?
  • When should the doctor/nurse encourage the patient to collect and provide data?
  • Is the patient's consent needed to process non-medical data they provided in the EHR?
  • How important is it to consider PGHD regarding communication with the patient, cooperation to reach therapeutic or preventive goals (compliance), and further encouraging the patient's engagement?
  • How to utilize the patient's active approach?
  • If the data collected by the patient has no clinical value, does it have any psychological value ("the patient feels the data is important")?
  • Which categories of data should always be analyzed?
  • How much time should be allocated to PGHD analysis so that other parts of the visit are not neglected?
  • How to discuss the data with the patient?
  • Should a guide be created to make it easier for patients to collect and provide data to the doctor? (educational materials)
  • Is it possible for a planned or existing patient e-portal/app to include PGHD?
The most important part of the process is identifying the patient's unmet needs, as they can lead to an increased interest in gathering data. The doctor can also ask the patient directly what symptoms made them decide to collect data. The research shows that PGHD brings more benefits than harm and that doctors should encourage their patients to collect data. But such an open approach should be backed by the right strategy so that doctors do not feel overwhelmed when they have to check data delivered by the patients. It will take a long time until PGHD become integrated with the medical data stored in the EHR.