15 Scenarios for large data applications in the healthcare industry

Gartner presented the BI concept in 1989. Gartner further upgraded the BI concept to advanced analysis (Advanced Analytics) in 2008. 2011, McKinsey explained the concept of large data. The names are different, but the problems they are trying to solve have never changed. However, large data analysis techniques now deal with more massive, diverse, real-time (Volume, produced, Velocity) data than 20 years ago, that is, large data. Compared with bi 20 years ago, large data analysis is now able to generate greater commercial value, and the development of large data storage and analysis technology has benefited from the proliferation of data in business scenarios and the diversification of data types.

So before implementing a large data analysis project, the enterprise should not only know what technology to use, but should know when and where to use it. In addition to the Internet companies that started using big data earlier, the healthcare industry may be one of the traditional industries that make big data analysis the first to flourish. The medical industry has long encountered the challenges of massive data and unstructured data, and in recent years many countries have been actively promoting the development of medical information, which makes many medical institutions have the funds to large data analysis. Therefore, the medical industry will be with the banking, telecommunications, insurance and other industries to enter the big Data era. In its report, McKinsey pointed out that the removal of institutional barriers, large data analysis can help the U.S. health services in the year to create 300 billion of dollars in value-added. This article lists 15 applications in 5 major areas of medical services (clinical business, payment/pricing, research and development, new business model, public health), where the analysis and application of large data will play a significant role in improving medical efficiency and medical outcomes.

Clinical Operation

In clinical practice, there are 5 major scenarios for large data applications. McKinsey estimates that if these applications are fully used, the nation's health-care spending will be reduced by 16.5 billion dollars a year alone.

1. Comparative effect Study

By comprehensively analyzing the patient's characteristic data and curative data, and comparing the effectiveness of various interventions, we can find the best treatment pathway for specific patients.

Research based on efficacy includes comparative effects studies (comparative effectiveness Research, CER). The research shows that for the same patient, the medical service provider is different, the medical care method and the effect are different, the cost also has the very big difference. Accurate analysis of large data sets, including patient signs data, cost data, and curative data, can help physicians determine the most effective and cost-effective treatment in the clinic. The implementation of CERs in the medical care system will likely reduce excessive treatment (such as avoiding the treatment of those side effects that are significantly more effective), as well as inadequate treatment. In the long run, both excessive treatment and inadequate treatment will have a negative impact on the patient's health and lead to higher medical costs.

Many medical institutions around the world (such as the United Kingdom Nice, Germany IQWiG, the Canadian General Drug Inspection Agency, etc.) have begun the CER project and have achieved initial success. The Recovery and Reinvestment Act, adopted by the United States in 2009, is the first step in that direction. Under this Act, the Federal Coordinating Committee for Comparative Effects was established to coordinate the study of the comparative effects of the entire federal government and to allocate 400 million of dollars in funding. There are a number of potential problems to be addressed in this effort, such as the consistency of clinical data and insurance data, and the current lack of EHR standards and interoperability requires a widespread and hasty deployment of an EHR that may result in difficult integration of different datasets. Again, patient privacy, to protect the patient's privacy, but also provide enough detailed data to ensure the effectiveness of the analysis results is not an easy thing. There are also institutional issues, such as the current United States law prohibiting medical insurance institutions and Medicaid service centres (Centers for Medicare and Medicaid services) from using cost/benefit ratios to make reimbursement decisions, So even if they find a better way through big data analysis it's hard to implement.

2. Clinical Decision Support System

The clinical decision support system can improve the work efficiency and the quality of diagnosis and treatment. The current clinical decision support system analyzes the entries that doctors enter, comparing them with medical guidelines, thus reminding physicians to prevent potential errors, such as adverse drug reactions. By deploying these systems, medical service providers can reduce the number of medical accidents and claims, especially those caused by clinical errors. Within two months of the study of the Metropolitan Pediatric Intensive Care Unit in the United States, the clinical decision support system reduced the number of adverse drug events by 40%.

The large data analysis technology will make the clinical decision support system more intelligent, which benefits from the increasing ability to analyze the unstructured data. For example, image analysis and recognition techniques can be used to identify medical imaging (X-rays, CT, MRI) data, or to tap medical literature data to establish a medical expert database (as IBM Watson did) to advise physicians on their treatment. In addition, the clinical decision support system can make most of the workflow flow flow to the nursing staff and assistant doctors, so that doctors from the time-consuming simple counseling work freed up, so as to improve treatment efficiency.

3. Medical Data Transparency

Improving the transparency of medical process data can make the performance of medical practitioners and medical institutions more transparent and indirectly promote the improvement of medical service quality.

Based on the set of operational and performance data sets for medical service providers, data analysis can be performed and visual flowcharts and dashboards created to facilitate transparent information. The goal of the flowchart is to identify and analyze clinical variations and sources of medical waste, and then optimize the process. Only the release of cost, quality and performance data, even without the corresponding material rewards, can often promote performance improvement, so that medical service providers to provide better services, and thus more competitive.

Data analysis can lead to streamlining business processes, reducing costs through lean production, finding more efficient employees that meet the needs, thereby improving the quality of care and providing a better experience for patients, as well as providing additional performance growth potential for healthcare providers. The U.S. medical Insurance and Medicaid Services Center is testing the dashboard as part of a proactive, transparent, open, collaborative government. In the same spirit, the U.S. Centers for Disease Control and Prevention (Centers for Disease controls and Prevention) has released medical data, including business data, publicly.

Public release of medical quality and performance data can also help patients make smarter health care decisions, which will also help healthcare providers improve overall performance and become more competitive.

4. Remote Patient Monitoring

Data are collected from the remote monitoring system for chronic patients, and the analysis results are fed back to the monitoring equipment (see if the patient is complying with the doctor's advice) to determine future medication and treatment options.

In the 2010, 150 million chronically ill people in the United States, such as diabetes, congestive heart failure, and high blood pressure, accounted for 80% of the health care costs of the health care system. The tele-patient monitoring system is very useful for treating patients with chronic diseases. The tele-patient monitoring system includes a household heart monitoring device, a blood glucose meter, and even a chip pill, which is transmitted to the electronic medical records database in real time after the chip pill is ingested by the patient. For example, remote monitoring can alert physicians to timely treatment for patients with congestive heart failure to prevent emergency situations, as one sign of congestive heart failure is the weight gain caused by water retention, which can be prevented by remote monitoring. The more benefit is that, through the analysis of the data produced by the remote monitoring system, the patient's hospitalization time can be reduced, the emergency volume reduced, and the target of increasing the ratio of family care and outpatient appointment should be achieved.

5. Advanced analysis of patient files

Applying advanced analysis in patient files can determine who is susceptible to a particular type of disease. For example, applying advanced analysis can help identify patients who are at high risk of developing diabetes and allow them to receive preventive health care programmes as early as possible. These methods can also help patients find the best treatment options from existing disease management programs.

Payment/Pricing

For medical payers, medical services can be priced better through large data analysis. In the United States, for example, this will have the potential to create 50 billion dollars a year worth, half of which comes from a reduction in national health spending.

1. Automation System

Automated systems such as machine learning technology detect fraudulent behavior. Industry assessment, the annual 2%~4% medical claims are fraudulent or unreasonable, so detection of claims fraud has great economic significance. A comprehensive and consistent claim database and corresponding algorithm can detect the accuracy of claims and detect fraudulent behavior. This fraud detection can be retrospective or real-time. In real-time detection, automated systems can identify fraud before payment occurs and avoid significant losses.

2. Pricing schemes based on health economics and curative effects

In drug pricing, pharmaceutical companies can participate in sharing treatment risks, such as setting pricing strategies based on therapeutic effects. The benefits to medical payers are clear and help control health-care costs. For patients, the benefits are more straightforward. They are able to get innovative drugs at reasonable prices, and these drugs are based on therapeutic research. For pharmaceutical companies, better pricing strategies are also good. They have access to higher market access possibilities, and can also gain higher incomes through innovative pricing schemes and more targeted therapeutic drugs.

In Europe, there are now pilot drug pricing projects based on health economics and efficacy.

Some medical payers are using data analysis to measure the services of medical providers and to price them based on service levels. Medical service payers can pay based on medical effects, and they can negotiate with health care providers to see if the services provided by the medical service provider meet a specific benchmark.

Research and Development

Medical products companies can use large data to improve research and development efficiency. Take the United States as an example, which will create more than 100 billion dollars a year worth.

1. Prediction Modeling

Pharmaceutical companies in the development phase of new drugs, through data modeling and analysis, to determine the most efficient input-output ratio, and thus equipped with the best resource mix. The model is based on the data set in the early stage of the clinical trial of the drug, predicting clinical outcomes as timely as possible. Evaluation factors include product safety, effectiveness, potential side effects and overall test results. Predictive modeling can reduce the cost of research and development of pharmaceutical products companies, and after data modeling and analysis to predict drug clinical results, it is possible to postpone the study of suboptimal drugs or to stop expensive clinical trials on suboptimal drugs.

In addition to the cost of research and development, pharmaceutical companies can get a quicker return. Through data modeling and analysis, pharmaceutical companies can bring drugs to market faster, produce more targeted drugs, and have higher potential market returns and treatment success rates. The usual new drugs from research and development to market time of about 13, the use of predictive models can help pharmaceutical companies 3-5 years earlier to bring new drugs to market.

2. Statistical tools and algorithms for improving clinical trial design

The use of statistical tools and algorithms can improve the level of clinical trial design and more easily recruit patients during clinical trials. By digging up the patient data, the patients were evaluated to see if they met the test conditions, thus speeding up the clinical trial process, proposing more effective clinical trial design and finding the most suitable clinical trial base. For example, a trial base with a large number of potentially eligible clinical trials may be more desirable, or a balance between the size and characteristics of the test patient population.

3. Analysis of clinical experimental data

Analysis of clinical trial data and patient records can be used to determine more indications of drug and to detect side effects. After analysis of clinical trial data and patient records, the drug could be repositioned or the marketing for other indications could be achieved. Timely or near-real-time collection of adverse reaction reports can promote drug vigilance (drug alert is the safety and security system of listed drugs, monitoring, evaluation and prevention of adverse drug reactions). Or, in some cases, clinical trials suggest some cases but not enough statistical data to prove that the analysis based on clinical trial data can now provide evidence.

These analysis projects are very important. It can be seen that the number of drug withdrawals in recent years has hit new highs, and the withdrawal of drugs may have devastating consequences for pharmaceutical companies. The Vioxx, which was removed from the market in 2004, caused 7 billion of billions of dollars in losses to Merck, resulting in a 33% loss of shareholder value in just a few days.

4. Individualized Treatment

Another big data innovation that is promising in research and development is to develop personalized therapies through the analysis of large datasets such as genomic data. This application examines genetic variability, susceptibility to specific diseases and the response to special drugs, and then considers individual genetic variants in drug development and medication.

Personalized medicine can improve health care effects, such as providing early detection and diagnosis before a patient has symptoms. In many cases, patients use the same treatment regimen but the effect is different, in part because of genetic variation. For different patients to take different treatment programs, or according to the actual situation of patients to adjust the dosage of drugs, can reduce side effects.

Personalized medical care is still at an early stage. In some cases, McKinsey estimates, reducing prescription doses can reduce the cost of 30%~70% medical care. Early detection and treatment, for example, could significantly reduce the burden on health systems caused by lung cancer, since early surgery costs are half the cost of late treatment.

5. Analysis of disease patterns

By analyzing the patterns and trends of disease, it can help medical products enterprises to make strategic research and development investment decision, to help them optimize their research and development priorities and to optimize the allocation of resources.

New business model

Large data analysis can bring new business model to medical service industry.

1. Summarize clinical records of patients and medical insurance data sets

Summarizing the clinical records and medical insurance data sets of patients and conducting advanced analysis will improve the decision-making ability of medical payers, medical service providers and pharmaceutical enterprises. For example, for pharmaceutical companies, they can not only produce better therapeutic drugs, but also ensure that medicines are marketable. The market for clinical records and medical insurance data sets is just beginning to develop, and the pace of expansion will depend on the speed with which the health care industry completes EMR and evidence-based medicine.

2. Network Platform and community

Another potential big data startup business model is the network platform and large data, which have produced a lot of valuable data. For example, patientslikeme.com website, patients can share treatment experience on this site; Sermo.com website, doctors can share medical insights on this site; Participatorymedicine.org website, a non-profit organization that operates a website that encourages patients to be actively treated. These platforms can be valuable sources of data. For example, sermo.com fees to pharmaceutical companies, allowing them to access member information and online interactive information.

Public Health

    The use of large data can improve public health monitoring. The public health sector can quickly detect infectious diseases by covering the country's patient electronic medical records database, conduct comprehensive surveillance and respond quickly through integrated disease surveillance and response procedures. This will have many benefits, including reduced medical claims, lower infection rates and faster detection of new infections and outbreaks in the health sector. By providing accurate and timely public health advice, public health risk awareness will be greatly enhanced, while reducing the risk of infectious disease infection. All this will help people to create a better life.