Three categories of patient information seem to be relevant at first glance:
- Information about me. Patients are more and more interested to know about their situation and what can be done about it. This is their personal view of the medical data and a reasonable abstract of what healthcare delivery is doing on it. It is just impossible to tell what amount of information is just enough for an individual patient. However it is possible to put the patient in a position to approximate this situation. Access to libraries, information packets and resource centers as offered in the Planetree model [Frampton] supplemented by Electronic Records accessible to patients, eLearing modules and community interaction provide valuable possibilities to improve the personal level of literacy.
- Information about the healthcare provider. Whether family physician or primary care hospital patients need to know what healthcare providers are doing about their situation. This is everything from the display of general hospital information, patient pathways up to the possibility to schedule admission or examination.
-
Information about others. This information is needed to build sustainable informational and personal relationships by the patient. The above mentioned communities whether virtual or nor are excellent possibilities to study what is happening to other patients and how they are treated in comparable situations. But also information about family members and relatives is necessary to support patients.
Having access to this information is good but having them separate is not very helpful. The real added value is to integrate all relevant patient information into an information chain that represents the changing informational need of the patient. Integration of information usually deals with provider centric information models. Standards in his domain are HL7, Snomed, ICD etc. They are of little relevance when it comes to patient centered scenarios. These standards are dedicated to information exchange within knowledge intense scenarios making heavy use of modeling relationships in medicine and healthcare delivery. Patient situations nee more lightweight integration like tagging, rating and commenting. Lightweight means here that the amount of information needed from the patient to integrate or link information is rather small. So are the interactions performed by the patient. No heavy weight diagnostic procedures to elaborate relevance by an expert system. Spontaneous combinations of information exchange take place like watching a video about a hospital after searching information about hospitals that deliver hip operations afterwards a discussion with relatives or other patients makes sense. The point here is not that this is just possible or reasonable, it should be done and documented in a timely fashion that leverages a simple form of causality. “I’m searching information about hip operation which leads me to a video about hospital XY and shows me things I really need to discuss with someone”. This chain of acts or events is important for the patients understanding but nevertheless ignored by standard healthcare delivery. Collecting these acts of informational behavior is important for companies like Amazon or Google because the customer is their most valued target. These companies compile these interactions into recommendations like ‘ Customers who bought this also bought that …’. This is not the case in healthcare delivery though it should be to step away from managing just illness instead of proactively anticipating patient needs and planning patient pathways. I think it’s worth to think a moment about what really can be done for patients stepping away from reaction in to anticipation. it’s not just easy. Health is not a product with fixed attributes. The equivalent product for patients is information. Within the three categories mentioned recommendation makes sense. ‘Patients with symptoms … searching for … also searched for …’. Having such a possibility together with collaboration with physicians and patients is far better than nothing. I would still warn against recommendation of products and drugs without guidance, but experience of others and knowledge is a reasonable target for recommendation. A hardly know any patient centered website implementing this but I’d be grateful for any hint.
Conclusion
Collecting information from and for patients is not trivial. Informational need changes according to the situation and has to be reassessed constantly. Integration of various sources or categories of information is useful if an added value is delivered. Collective Intelligence and recommendation of related cases, information and patients may be worth a try. The difficulty is that we are not dealing with products but with patients and their needs. It may be very difficult to assess what and how much information is good for a patient in a given situation. However a combination of content-based and collaboration-based intelligence to determine what are the intentions of the patients in relation to the information requested. For healthcare delivery is important to understand that information is an important part of the healthcare delivery process and that it has to be customized just like any other therapeutic intervention to the patients needs. Additionally information collected from patient interactions may be channeled into epidemiologic research [Public Health 2.0].
Patterns
content-based recommendation
collaboration based recommendation
Links
- GILL PINNER, MRCPsych, 2000, Truth-telling and the diagnosis of dementia, The British Journal of Psychiatry (2000) 176: 514-515
- Susan B Frampton (Editor), Patrick A. Charmel (Editor), Planetree (Editor), Putting Patients First: Best Practices in Patient-Centered Care, 2nd Edition, Wiley
- Dr. James F. Drane, Profesor Emeritus,University of Edinboro Pennsylvania, Honesty in Medicine: Should Doctors Tell the Truth?
- Public Health 2.0
