The What and Why of Big Data in Healthcare
Gary Monk writing on his blog takes look at the term big data and its application in healthcare. Big data has been around for since the 19th century. Monk references Commodore Maury’s use of ship logbooks to identify efficient sea routes by . He notes that this was a secondary purpose of the original intent.
There are reasons why big data is BIG today:
- Ease of collection
- “The Data Exhaust” This is the by-product of our digital life, phone calls, searches, Tweets, etc.
- Processing Power
- Mindset We are getting more skilled at using this data
Monk references the following examples
- Tracking and predicting epidemics: Google flu
- Medication Adherence He links to “What is the FICO Medication Adherence Score?” FICO now measures how well we will adhere to a prescription
- Human health data
Monk ends his article with the simple message we need to find better and newer ways to integrate the data we are collecting in healthcare to deliver knowledge, information, and findings that can change outcomes.
Just to note here is some secondary use of big data in an excellent healthcare study “Using Twitter to Examine Smoking Bhavior and Perceptions of Emerging Tobacco Products”
People with Doctors Interested in EMRs, Where is the Easy Button?
HealthPopuli share some data from the EMR Impact Survey. One in two insured people with a regular physician want to sample EMR but need a HCP to suggest it. Oh and needs it to be easy to use.
The Impact Survey looked at 1,000 US online consumers age 25 to 55 with insurance and have seen their physician in the past three years. One in four have accessed their EMR online the survey. The survey further breaks down the data. 18% of this group have never tired or are interested in EMR, 9% have tried EMR once or twice and most are women, 13% are considered regular users. This group emails their physician and 1/3 are caregivers. I would say this group are ePatents.
The most interesting group are the 52% who are identified as Interested Non Users. They are ready to adopt but are less satisfied with their physicians and would switch to a physician using EMR in their practice. One half of this drop has physicians using EMR but 42% don’t know where or how to access it.
Barriers to use solo or private practice, rural locations, HCP’s who were not encouraging, and the perception that they are difficult to use especially by those who are not using them.
That final point sounds like a business idea or an educational initiative: teach consumers to use EMR’s.
Blood pressure management is the holy grail of adherence and management with only one half of patients with hypertension have their blood pressure controlled. Thiboutot, Sciamanna, et. al. have designed a study to test the effect of an intervention which would help patients ask questions at the point of care in order to encourage PCP to appropriately intensify blood pressure treatment.
500 patents and their PCPs were recruited and randomized into two groups.
(1) intervention condition in which patients used a fully automated website each month to receive tailored messages suggesting questions to ask their PCP to improve blood pressure control, and (2) control condition in which a similar tool suggested questions to ask about preventive services (eg, cancer screening). The Web-based tool was designed to be used during each of the 12 study months and before scheduled visits with PCPs. The primary outcome was the percentage of patients in both conditions with controlled blood pressure.
The conclusion: interactive web site designed to overcome clinical inertia for HTN care did not lead to improvements in blood pressure control. In addition participant adherence to the intervention was high, the control intervention saw positive changes in the use of preventive services and it lead to more discussion of HTN relevant tests. Participants were more likely to discuss questions with their PCP. BUT these discussion didn’t lead to improvements in blood pressure control.
I found the hypotheses explaing why no effect on HTN was observed was interesting tells us a great deal that should be considered if we are to continue this study.
There are several possible hypotheses to explain why no effect on blood pressure was observed. First, patients may not have been comfortable asking for intensifications to their medication treatment plan because of a concern about questioning the expertise of the provider.
Although Kravitz and colleagues  observed that prompting standardized patients to ask for a depression treatment increased the chances of receiving treatment, we believe that asking for a drug intensification for hypertension is a very different act. In a clinical encounter for depression, for example, the patient possesses more data than the provider upon which decisions will be made (eg, depressive symptoms). In a clinical encounter for hypertension, this is reversed; the provider typically has more data than the patient (eg, blood pressure values). Because of this difference, encouraging patients to ask questions about their blood pressure control has inherent limitations, making this intervention strategy questionable for this setting.
Asking for medication intensification may have been perceived by patients as questioning the judgment of the provider, which may have created a barrier to asking for medication intensifications. This is consistent with the observation, in that the intervention led to more conversations about testing for creatinine and urine protein, but no differences in conversations about intensifying medications.