• EHR
  • IOp's now measured in a better way

    Twelve EHR providers have agreed to a new way to measure InterOperability (IOp). KLAS Research has announced the agreement by twelve EHR providers at the KLAS Keystone Summit in Utah, USA, to help solve essential IOp challenges. Now, it’ll be measured objectively. The KLAS announcement doesn’t say what the measures are. FierceHealthIT reports that they’re a combination of:

    Transaction counting The experience of clinicians.

    The twelve vendors are:

    Allscripts athenahealth Cerner eClinicalWorks Epic GE Healthcare Greenway Healthland McKesson MEDITECH MEDHOST NextGen Healthcare.

    The next step’s to put a cohesive plan in place to launch and monitor the measurement. After that, the task is to implement it. From these activities, Africa’s health systems can ask to see the performance of the vendors who bid for their EHR service. It’s reasonable for Africa’s health systems to include the availability of the measurement information as part of their vendor accreditations.

  • Patient portals are effective

    Should patients have access to their medical records? It’s an important decision for Africa’s eHealth if it increases the benefits of EHRs. There may be trend emerging.

    In 2012, a study in the Journal of Internet Medical Research (JMIR) of patient portals’ effectiveness by and Austrian team found:

    Five papers presenting found studies No statistically significant changes between intervention and control group in the two randomised controlled trials investigating the effect of patient portals on health outcomes Significant changes in the patient portal group, compared to a control group for: Quicker decrease in office visit rates Slower increase in telephone contacts Increase in the number of messages sent Changes of the medication regimen Better adherence to treatment.

    A USA study in 2015, also reported in the JMIR, found the availability of notes following visits to GPs was associated with improved adherence by patients to prescribed antihypertensive medications, used to treat high blood pressure. It didn’t find an improvement for antihyperlipidemic medications, used as cholesterol busters.

    Surveys, interviews, and focus groups have shown that patients taking medications and offered portal access to their GPs notes reported better adherence to their regimens. There’s a big but. The study says that objective confirmation hasn’t been reported, so the team completed a retrospective comparative analysis at a site of the OpenNotes quasi-experimental trial. Quasi-experimental research has similarities with the randomised controlled trials, but they don’t have a random assignment to treatment or control.

    Patients in Danville, Pennsylvania, older than 17, in the Geisinger Health Plan insurance, and taking at least one antihypertensive or antihyperlipidemic agent from March 2009 to June 2011, provided data for the study. From 2010, patients were invited and reminded to use the portal to read their GPs notes. Control patients had portal access too, but their GPs notes weren’t available. There were 2,147 patients, 756, 35%, in the intervention group, and 1,391, 65%, in the control group.

    The 756 patients with access to their GPs notes were more compliant to antihypertensive medications regimens. The rates were nearly 80% compared to 75% for the control group. For antihyperlipidemic medication, the rates were about 77% for both groups.

    The study says as the “Use of fully transparent records spreads, patients invited to read their clinicians’ notes may modify their behaviours in clinically valuable ways.” How can Africa’s eHealth take advantage of it?

  • There are five main IOp barriers

    If interoperability (IOp) is critical to successful eHealth, why hasn’t it been fixed after all this time? The Government Accountability Office (GAO) in the USA’s report ELECTRONIC HEALTH RECORDS Nonfederal Efforts to Help Achieve Health Information Interoperability says there are five big barriers to EHR IOp. They seem to fit most countries eHealth endeavours The five are:

    Insufficiencies in standards for EHR interoperability Variation in state privacy rules Accurately matching patients’ health records Costs Need for governance and trust among entities.

    The GAO’s view of IOp for EHRs has five components: 

    View results from diagnostic procedures conducted by other providers to avoid duplication Evaluate test results and treatment outcomes over time regardless of where the care was provided to better understand patients’ medical histories Share a basic set of patient information with specialists during referrals and receive updated information after the patient’s visit with the specialist to improve care coordination View complete medication lists to reduce the chance of duplicate therapy, drug interactions, medication abuse, and other adverse drug events Identify important information, such as allergies or pre-existing conditions, for unfamiliar patients during emergency treatment to reduce the risk of adverse events.

    The review was need for many reasons:

    EHR IOp’s seen by many healthcare stakeholders as a requirement for improving healthcare It’s remained limited, despite the federal government’s role in guiding IOP ahead Many initiatives needed for IOp are pursued by non-federal stakeholders who have to develop and implement infrastructure needed for national IOp.

    The overall finding is that IOp is still a Work-In-Progress (WIP). Removing the barriers, sometimes called inhibitors or challenges, is always a priority for eHealth. Acfee‘s report Advancing eHealth in Africa found loads of challenges for Africa’s health systems, fat outweighing other eHealth themes. IOp’s only one of them. That it’s a WIP isn’t too surprising. eHealth, like life, "Is a journey, not a destination," as the poet Ralph Waldo Emerson said. It’s much more elegant than saying it’s a WIP.

  • EHRs and HIE can help forecast demand

    When the USA’s State of Maine set up HealthInfoNet, its Health Information Exchange (HIE), it integrated predictive analytics into it with an online population risk surveillance dashboard. It allows researchers to conduct real-time surveillance for field staff and population health managers. It seems that it’s now paying off. A report in the Journal of Medical Internet Research (JMIR) says researchers have used the HIE to create an online risk model using EMR data to predict the healthcare resources patients need for the next six months.

    Some patients are the equivalent of frequent flyers. They’re a relatively small proportion of patients, but use a relatively large amount of healthcare resources. They have a specific focus in the risk model that focuses resources on them, and treating them separately.

    The model’s been tested for reliability at individual patient level and population level. It aggregated data from more than a million patients, and extended back over the previous year.

    Maine’s HIE initiative confirms a strategic goal for EMRs and EHRs in Africa. Initially, many benefits were seen to derive from sharing data between healthcare professionals. This’s still an important benefit, but now, using the data to forecast short-term demands and plan the availability and utilisation of healthcare resources steps up the benefits. Africa’s health systems can now include this, and the resources it needs in their eHealth plans.

  • Is Africa's eHealth usability at risk?

    As Africa moves ahead on using EHRs, regulation, usability and procurement will converge. Usability’s an essential component of benefits, and regulation’s essential to setting EHRs’ and other eHealth usability standards and supplier certification. Procurement’s where these are applied. Africa’s health systems have a few challenges.

    Their eHealth regulations are well behind good practice. A study of eHealth regulation in Sub-Saharan African countries by South Africa’s Greenfield Management Solutions found that specific eHealth regulations are minimal, with reliance on general regulations such as data protection and telecommunications regulations. This status provides no basis for setting User-Centred Design (UCD) standards for usability for eHealth vendors to apply. Two main sources for these are one generic and one for EHRs.

    ISO 9241-210:2010 is generic. It has six core design principles for usability where design’s:

    Based on an explicit understanding of users, tasks and environments Users are involved throughout design and development Driven and refined by user-centred evaluation An iterative process It addresses the users’ whole experience Dealt with by multidisciplinary teams with an appropriate range of skills and perspectives.

    The USA has specific requirements. It’s National Institute of Standards and Technology has adopts NISTIR 7741 NIST Guide to the Processes Approach for Improving the Usability of Electronic Health Records. It’s principles are:

    Understand user needs, workflows and work environments Engage users early and often Set user performance objectives Design the user interface from known human behaviour principles and familiar user interface models Conduct usability tests to measure how well the interface meets user needs Adapt the design and iteratively test with users until performance objectives are met.

    It has six types of compliance:

    0 – Incomplete, so unable to carry out the process

    1 – Performed, where individuals carry out the process

    2 – Managed, where quality, time and resource requirements for the process is known and controlled

    3 – Established, where the process is carried out as specified by an organisation, and resources are defined

    4 – Predictable, where the performance of the process is within predicted resource and quality limits

    5 – Optimising, where an organisation can reliably tailor the process to particular requirements.

    By including these usability standards in eHealth regulations, health systems can expect vendors to comply with good practice, leading to good benefit’s realisation. Health systems can also check how far vendors comply with the standards. This is the crunchy bit.

    A team, led by Dr Raj M. Ratwani of MedStar Health, Washington, DC, has written a research letter to the Journal of the American Medical Association (JAMA). It says that many EHR vendors don’t comply with the usability standards. It’s an important finding because many EHRs have poor usability.

    The team reviewed documentation of 41 of the 50 certified vendors:

    34% failed to state their UCD process 46% used an industry UCD standard 15% used an internally developed UCD process 63% used less than the standard of 15 participants 22% percent used at least 15 participants with clinical backgrounds One vendor used no clinical participants 17% used no physician participants 5% used their own employees 12% lacked enough detail to determine whether physicians participated 51% didn’t provide the required demographic details.

    This poor compliance occurred in a highly regulated system, and was revealed. With African countries limited regulation, they’re more vulnerable to eHealth usability limitations, leading to reduced benefits and wasted resources. The next step’s simple: enhanced eHealth regulation for procurement. 

  • Microsoft says some EDBs for EMRs have holes

    EHRs have become a central pillar of eHealth. Across Africa, they’re steadily building up. 

    They also arouse some strong feelings about privacy, confidentiality and security. A study by Microsoft Research has shown that these anxieties are well-founded. The team of three, Muhammad Naveed from the University of Illinois at Urbana–Champaign (UIUC), Seny Kamara of Microsoft Research and Charles V. Wright from Portland State University, reviewed the “concrete security” of Encrypted Database (EDB) systems based on the design of CryptDB (SOSP 2011) and that rely on property-preserving encryption such as Deterministic Encryption (DTE) and Order-Preserving Encryption (OPE).

    They evaluated attacks empirically on EMRs, using real patient data from 200 US hospitals. “Alarming” was the finding. EDBs operating in a steady-state, with enough encryption layers peeled so the application can run its queries, allowed the recovery of a large amount of sensitive information. It allowed access to:

    80% of patient records from 95% of hospitals for attributes, such as age and disease severity 60% of patient records from over 60% of hospitals for DTE attributes such as sex, race and mortality.

     The study used four types of attacks:

    Frequency analysis, which decrypts DTE-encrypted columns `p-optimization, which decrypts DTE-encrypted columns Sorting attack, which decrypts OPE-encrypted columns Cumulative attack, which decrypts OPE-encrypted columns.

    While the study used EMR databases, it believes that its attacks would be as successful on many types of databases if appropriate auxiliary information is available. This raises security concerns for other healthcare databases such as registries. Just as concerning is that “All the attacks take less than a fraction of a second per hospital.”

    The study says that encryption methods used for the study shouldn’t be used for databases that store personal medical data. It’s not explicit on what these are. The cyber-security war goes on, and Africa needs to join.

  • How registries use EHRs for better healthcare

    Registries are seen as effective ways to improve health, develop proactive healthcare and manage population’s health. These are big claims. Achieving them needs rigorous research into how to assemble and use the data. The USA’s Agency for Healthcare Research and Quality (AHRQ) is investing in researching how to use eHealth, including registries to improve healthcare quality, safety, delivery and patients’ outcomes. These are goals for Africa’s health systems too.

    An AHRQ brief sets out how it’s helping clinicians to use Big Data to transform healthcare and develop personalised medicine. The initiative includes collaborations of informatics experts, researchers and clinicians to develop better ways to link EHRs with claims, pharmacy data and diagnostic data. It also looks at ways to collect and use specific, individual information about patients, such as medication side effects, pain after surgery, disease symptoms, and experiences of healthcare to create comprehensive pictures of patients’ experiences. Maintaining and protecting patients’ confidentiality is paramount. The project also supports the development of new analytic tools to use data promptly to produce information that improves decisions and clinical care.

    Linking data from patients, clinicians, and researchers across many clinical sites can create enhanced clinical registries. These aim to lay a foundation for a learning healthcare system by providing real-time access to knowledge about each patient’s healthcare. Clinicians can then use the knowledge to devise better ways to treat and care for patients. This is essential to achieve broad, proactive healthcare.

    Projects already underway deal with paediatric Inflammatory Bowel Disease (IBD), including Crohn's disease and ulcerative colitis. The condition can have many severe complications. AHRQ says clinicians are finding new answers from the Enhanced Registries Project. Since 2007, the results include:

    Children's remission rate increased from 55% to 78% 49% sustained remission for at least one year 94% achieved satisfactory growth 91% achieved satisfactory nutrition 95% don’t take steroids.

    Other projects include a hospital surgical checklist and a total joint replacement project. Both have their enhanced registries.

    These AHRQ initiatives show a way of maximising EHRs’ benefits, and the need for eHealth projects that are much wider than just ICT. With many of Africa’s eHealth programmes still at early stages, the time is right for them to start to put these registry research initiatives in train.

  • Trigger algorithms find cancer early

    Early diagnosis of cancer is known to improve the chances of recovery.  A study in the Journal of Clinical Oncology by a team from Baylor College of Medicine and the Michael E. DeBakey Veterans Affairs Medical Center tested whether prospective use of trigger algorithms from EHR data could identify patients at risk of diagnostic delays for cancer and could prevent delays in diagnostic evaluations.

    Trigger algorithms decide if data for observed events and interactions indicate specified phenomena needs some form of action. There’s not much room for error in their use. The team used them to search patients’ EHRs with two groups of primary care providers of 36 providers in each. The search was for initial screenings for colorectal and prostate cancers showed abnormal findings, but had no associated follow-up actions. It found that using trigger algorithms can reduce the time needed to evaluate diagnoses for colorectal and prostate cancer. It also improves follow ups for more patients. It proposes that similar interventions could improve timeliness of diagnoses of other serious conditions.

    The Cluster Randomised Controlled Trial (CRTC) compared times to diagnostic evaluation and proportions of patients followed up between intervention and control cohorts based on final review at seven months. Intervention steps included queries of the EHR repository for patients with abnormal findings and lack of associated follow-up actions, manual review of triggered records, and communication about these with the two groups.

    Of the 10,673 patients with abnormal findings, the trigger flagged 1,256, nearly 12%, as high risk for delayed diagnostic evaluation. Times to diagnostic evaluation were significantly lower. For colorectal cancer, intervention patients had a median average wait of 104 days compared to 200 days for control patients. For prostate, the time was 144 days compared to 192. For lung cancer, the average time difference of 65 to 93 days wasn’t significantly different.

    The result is that electronic trigger-based interventions seem to be effective in reducing the time to diagnostic evaluation of colorectal and prostate cancer. It also improves the proportion of patients who receive follow-up. Similar interventions could improve timeliness of diagnosis of other serious conditions, but for lung cancer, conventional approaches may be as effective as using trigger algorithms.

    As non-communicable diseases in Africa are on the rise, investing in EHRs and using the data for algorithms to improve care can offer more effective healthcare. eHealth strategies should include the development of this capacity.

  • Patients like using their EHR data

    EHRs offer an opportunity for patients to access their medical records. What’s the benefit for them? A study in the Joint Commission Journal on Quality and Patient Safety on the anticipated effects of Open Notes, a USA initiative to give patients access to their notes written by their health workers, says it offers potential to improve the healthcare. A team from Beth Israel Deaconess Medical Center  and Harvard Medical School provides examples that Africa’s health systems can aim to achieve.

    A press release in EurekAlert summarises the findings from patients as:

    Catching medication errors Remembering the next steps in their healthcare Better plan adherence, including medication reminders Better error reporting Improved co-ordination of care for informal caregivers of vulnerable patients with many providers and appointments Reduced diagnostic delay Safer care Reminders for follow-up appointments.

    These are examples, so indicate the potential, not the probability of cost-effectiveness or net benefits. The findings still indicate the gains of allowing patients to access their EHRs. Some EHRs in Africa already do this. HealthSpace, an EHR for GPs in South Africa is an example. Using mHealth for SMS reminders can help, but can’t easily replicate the full range of medical information in EHRs. The two services can run alongside each other.

  • Some USA physicians now like their EHRs

    Black Book Rankings, a USA polling firm, has found an important nuance on USA physicians’ dissatisfaction with EHRs. Virtual-Strategy has a report on its survey of large practices. It says the polling firm discovered that almost two thirds of practices with more than 25 clinicians commended the enhancements in functionality, service and innovation. In 2013, 92% were dissatisfied.

     A summary of the trends of changes to satisfaction is:

    Executives and physicians in the practices say the improvements due to:

    Client education at 42% Product bundling at 31% Marketing at 26%.

    About 18% of large practices and clinics are discussing or implementing replacements for their original EHR by the end of 2016.

    It’s not all rosy. Decreases in satisfaction by physicians in clinics were:

    Customer support at 85% Implementation and training at 77% Failed regional connectivity attempts at 76%.

    These findings are more optimistic than previous surveys reported on eHNA, both by Black Book and other pollsters. The most recent was from AmericanEHR and the American Medical Association (AMA).

    It may be that larger practices may be able to use EHRs to improve their teamwork. They may also have more shared resources to exploit EHRs more effectively. Whatever the cause, it’s a more cheerful outlook for EHRs in primary care.

    For African countries, the range of findings indicates that they should consider carefully how they structure their EHR strategies and plans. Selecting the more viable practices may offer a good start.