Electronic Health Record-Keeping (Ehrs) According to Jensen,

¶ … electronic health record-Keeping (EHRs)

According to Jensen, Jensen & Brunak (2012)'s article entitled "Mining electronic health records: Towards better research applications and clinical care," scientists have a potentially invaluable source of information at their fingertips that can improve human health -- the data yielding by analyzing the electronic records of patients. "Mining of electronic health records (EHRs) has the potential for establishing new patient-stratification principles and for revealing unknown disease correlation" (Jensen, Jensen & Brunak 2012). One of the most common complaints about clinical trials is their limited nature: their accuracy may be compromised by relatively small numbers, limited demographic profiles of participants, and the difficulties of longitudinal analysis, all of which EHRs can potentially remedy.

The downside of using EHRs is that it involves using "scattered" and "heterogeneous" data not specifically designed for the purposes of research (Jensen, Jensen & Brunak 2012). Still, using such information is still vitally important, as the awareness of how specific genetic profiles can affect the development of various diseases has grown (Jensen, Jensen & Brunak 2012). As more and more hospitals implement EHR, with the encouragement of the federal government, the amount of potential information from such records will increase, regarding correlations of disease with specific patient genetic profiles as well as with health behaviors. This will improve early detection. However, what is valuable for healthcare collectively is not necessarily salutary for individual patients, whose privacy must be protected. In America, the Health Insurance Portability and Accountability Act (HIPAA) is designed to provide patients with safeguards; yet "despite such precautions, re-identification has been shown on some occasions to be a genuine risk, especially when data on human DNA are involved, as even a relatively small set of markers can enable unique identification" (Jensen, Jensen & Brunak 2012). Regionally and internationally, a lack of compatibility between EHR systems is yet another obstacle. Still, although the return on investment (ROI) for the organization may not always be clear in financial terms because of the logistical obstacles in implementing EHR, in terms of knowledge of human health, it is undeniably beneficial (Vogel 2003).

However, according to Gardner (2013) in her article, "The healthcare approach to big data," many healthcare organizations remain suspicious of IT. This may be due to the fact that the current methods of mining clinical data enthused about by Jensen, Jensen and Brunak (2012) must be viewed with caution. Instead of complaining that healthcare organizations have failed to value IT sufficiently compared with other industries, it must be asked why this is the case. In the drive for standardization, a great deal of valuable clinical insight may be lost from qualitative observations of clinicians scrawled on charts. "Big Data proponents now say it isn't necessary-or even desirable-to take structured data too much farther than it's already gone" (Gardner 2013). Also, not all sources of healthcare data are created equal. Data from healthcare claims, for example, is notoriously unreliable and lacking in the specifications about individual patient conditions from which it might be useful to draw meaningful hypotheses. And a final, 'incomplete' aspect of Big Data is how much is 'lost' regarding minute-by-minute information about patient vitals which is not stored the same way more permanent, final data may be: "One of our researches wanted to know if we could get the 'every minute' look at the data, and whether he could compare it to the medication administration record to see when a drug was given and what the effects were," but this information was not available (Gardener 2013). Gardner's article supports the notion that a full and integrated implementation of EHR is necessary but the methods through which data is accumulated must be refined for it to realize its promise and to justify its expense to organizations (Gardner 2013; Vogel 2003; Jensen, Jensen & Brunak 2012).