Continue in 2 seconds

Data Warehousing in the Health Care Industry - Three Perspectives

By
  • Carla McEachern, Les Stern, Lawrence Bell
Published
  • March 01 1998, 1:00am EST
More in

Health Care & the Data Mart:
Managed care learns to manage its customers

By Les Stern and Carla McEachern

With health care moving from a wholesale to a retail environment, health care marketers across the country have a new focus on customers. Like other industries, health care is beginning to realize the advantage of "knowing" and managing customers, particularly when trying to keep them. In the volatile health care industry, this customer knowledge is strategic for survival, best buoyed by the use of a marketing data mart. Adding analytical capabilities to the customer information that organizations have traditionally maintained, marketing data marts offer health care marketers greater speed and functionality not only to know their customers, but to hone their strategic planning.

Because of the competitive climate created by managed care, health care organizations (including providers such as health plans, hospital systems, physician management companies, etc.) can no longer rely on employer contracts, or the laws of geographic convenience, to win and retain customers. A marketing data mart provides in-depth customer knowledge, enabling a health care organization to lay the foundation for a relationship that makes members feel connected, more satisfied with overall services and less inclined to leave.

Strong marketing strategies driven by customer intelligence enable health care organizations to better understand and meet the needs of customers, maximize campaign results, build loyalty and, ultimately, improve bottom-line results. Today's customer-centric marketing practices utilize sophisticated data marts that drive intelligent marketing strategies--from the promotions mailed to the services offered. Building this technology-enabled strategy benefits marketing departments as well as the entire health care organization's strategic planning.

With a marketing data mart, health care organizations will be able to:

Understand Customer Value: By using the marketing data mart to acquire and organize strategic knowledge about customer value, the health care marketer can develop profiles of loyal customers--and determine customer lifetime value--using information such as demographics, behavioral or lifestyle data, product and service utilization, survey data and information about customer health care attitudes. The health care organization can use this knowledge to appropriately deliver care to its customers, keep its current customers happy, move other customers to higher levels of satisfaction and acquire new members that have the same profile as their highly satisfied customers.

Evaluate Products and Services: The marketing data mart provides information about the products and services customers use most often, which they most value and the time that elapses between their visits to the health care organization. Because customers in different segments tend to demonstrate different product and service affinities, marketers gain clues about which services to target to specific populations. The design and evaluation of services is always changing, and the health care marketers eventually identify certain packages of services that best attract new customers and that build loyalty among other customers.

Develop Contact Strategy: After the health care organization has selected prospects that closely resemble its best customers and evaluated the products and services being offered, it can better develop the contact strategies that will deliver the highest value for members. Using the marketing data mart, the health care organization can manage campaigns and track responses so that the contact strategy can be evaluated and refined over time. The marketers can test different media--from direct mail and newsletters to phone and Internet--and identify various ways to deliver value to both current and potential customers. Sachs Group's consumer research shows that an "engaged" customer--one who has used a service or had some kind of contact with the health care organization-- is more satisfied and more loyal. Ultimately, the best way to engage consumers is to know their needs and target marketing efforts in areas suited to them.

Strategic Planning: Strategic planning evolves when marketers "close the loop" and manage marketing programs on an ongoing basis. Using the marketing data mart to benchmark, track campaigns, analyze results and plan for the future based on their findings, marketers build a personalized strategic approach that takes health care marketing directly to the consumer. Health care organizations can then continually measure and refine the contact strategies that meet overall strategic goals.

While the marketing data mart enables health care organizations to implement customer-centric marketing, it also enables the organization to provide more of that customer information to everyone in the organization. By sharing the marketing-based information with medical managers, operations executives, sales, product managers, customer service representatives, etc., the health care organization can improve strategic as well as tactical planning. Anyone accessing the database throughout the organization can see and understand a customer's value to the organization as well as the products and services that may be of interest to the customers, allowing them to make appropriate up-sell and cross-sell suggestions. With everyone in the organization better equipped to understand and deliver customer value, real strategic planning emerges.

The marketing data mart equips health care organizations with knowledge about the value delivered to customers, along with what, how and when they use health care services--which is very powerful and actionable information. Over time, the health care organization better defines its most valuable customers and services, conserving resources while favorably serving customers with the services they want and need. In building these impressions, the health care organization yields significant and sustainable growth while retaining a more formidable core of loyal customers. By expanding the role of the marketing data mart to drive marketing strategies, health care marketers can develop customized contact strategies that focus on customers' individual needs, respond to management directives and provide absolute accountability.

The Pharmecutical Manufacturer - Designing the Data Warehouse

By Lawrence Bell

Anyone who works for a pharmaceutical company will tell you without hesitation, "Sales and marketing within the pharmaceutical industry is unlike sales and marketing in any other industry." The big secret is that a sales representative of a pharmaceutical manufacturer does not actually sell anything. Within other industries that have a more traditional sales model, a sales representative knows immediately after a sales call whether or not a sale was made. The potential customer either places an order or sends the sales representative away without a sale. This traditional model does not exist within the pharmaceutical industry. Instead, it is the responsibility of the sales representative to provide information by discussing product characteristics and benefits to the physicians and potential prescribers. The goal of the sales representative is to influence prescribing habits, so that when a patient comes into the office with symptoms X, Y and Z, it's their drug that is prescribed. This indirectly results in the sale of products at the pharmacy.

In the pharmaceutical manufacturing industry, information is critical to the success of the sales presentation process, which requires a detailed knowledge of the product. Certainly a product database with product specifications, clinical trial and patient outcome results could be an invaluable tool to the sales representative. However, this detailed product information is only one dimension of information that a sales representative requires to be effective. Pharmaceutical sales representatives and managers look for information that can provide insight into the management of the sales process. They need to know current business conditions (measures of the business or facts); internal measurements (measured ratios comparing results to internal factors); external measurements (measured ratios comparing results to external factors); and why changes have occurred.

In order to provide this information to support managers and sales representatives, it is necessary to collect data about the sales process, put it in one place and build reports against it. In an environment where we understand the business need and analytic requirements, we can see that a data warehouse could provide end users with a single place to go in order to get the answers to their many questions. Providing end users with transparent and timely access to information will improve the productivity of the organization and provide management with a competitive advantage. The automation of the reporting process will provide the analysts with the opportunity to actually perform analysis rather than spend time building reports.

Building the Data Warehouse

The building of the data warehouse in the pharmaceutical industry is similar to the process that would be followed in other industries; however, there are certain key factors which need to be considered during the development cycle:

  • This industry has a tremendous dependency on third-party data. Companies such as IMS America Inc. and Walsh America/PMSI actually gather information directly from the pharmacies, doctors, hospitals and managed care plans about patient care and prescriptions and resell this information to pharmaceutical companies. Since most of the inventory manufactured is sold directly to large wholesalers, without this third-party information, pharmaceutical companies would have very little information on the sale of their products.
  • Timing of the various data sources also has a significant impact on the reporting process. Some of the external data sources aren't available until months after the transactions occur which needs to be taken into consideration when constructing the time dimension and in building management reports.
  • External data needs to be integrated within the data warehouse and typically needs to be scrubbed to cross-match this data with internally maintained data about customers and products. Data cleansing is a critical component of the data warehouse for a pharmaceutical company and usually requires the implementation of software designed for this function.
  • The user community has a much larger geographical dispersion than in most industries, with the majority of the users being mobile field representatives. This complicates the process of delivering information to these users significantly due to the remote nature of their network access.
  • There are certain regulatory reporting requirements that can be satisfied through careful design of the data warehouse.

The Dynamic Nature of the Industry

The pharmaceutical industry has one of the most dynamic and volatile environments in business. With the threat of legislation and constant "improvements" to the health care industry, the sales process and, therefore, the data are in a constant state of flux. The most critical lesson that can be learned from the pharmaceutical data warehouse model is the need for flexibility and quick implementation. By the time a multi-year project is implemented, the industry could have transformed into something completely different. The data warehouse should also be constructed with adequate flexibility so that when a new twist gets thrust upon the industry, the data warehouse can be modified to handle these changes instead of having to be rebuilt from scratch.

The process of building a pharmaceutical manufacturing data warehouse is a complex and iterative process that is clarified only by understanding the uniqueness of the industry. As in other industries, a successful data warehouse is based upon the business need of the end users that require access to information in order to increase their effectiveness and productivity. Although a painful process at times, the benefit of a carefully constructed data warehouse has proven itself to be well worth the investment.

A Little Loss of Privacy may be Good for Patients' Health

by Diana J. P. McKenzie

In today's world of electronic patient records and large networks of health care providers, vast databases of patient information can be easily assembled and studied. The increased availability of such information raises numerous concerns with respect to patient privacy. Many of these concerns relate to the danger of unauthorized access from outsiders, such as reporters or computer hackers. Fortunately, a number of security devices are available to limit these types of risks.

However, other concerns have been raised in connection with how these databases are used by those inside or affiliated with the health care network. In this context, the sacrifices to patient privacy must be measured in light of the benefits of accessing and studying the available data. In many instances, patients--and the public in general--will be better off if responsible use of patient data is permitted.

Benefits of Access

Patients can benefit in many ways from the increased accessibility of their medical information, whether that data is "individually identifiable" or not. "Individually identifiable health information" has been defined as information that identifies a patient or with respect to which there is a "reasonable basis to believe that the information can be used to identify the patient."1 Some examples of individually identifiable information include records containing the patient's name, social security number or photograph. Although this type of data is often more useful for data harvesting purposes, it is also much more sensitive and thus subject to greater limits on use.

The following are some examples of how patients may benefit from increased accessibility to their individually identifiable medical records:

  • Researchers have identified a number of patients within a database that are statistically at risk of contracting a particular disease. If they can contact the patients' physicians, preventive measures can be taken to minimize the risk or control the condition.
  • An out-of-state patient is rushed to the emergency room, unconscious after an accident. If the emergency room physician is able to quickly access the patient's medical records and review his past medical history, the outcome of the treatment could be significantly improved.
  • A diabetic patient is on an extensive drug regime. If the pharmacist is able to access the patient's file, he would be able to offer a number of beneficial services. For example, he could detect potential allergic reactions to a prescription or adverse drug interactions, send reminders of prescriptions that need to be refilled, or notify the patient should a more affordable generic medication become available.
  • A patient is awaiting an organ transplant. Access to the patient database could facilitate locating the patient and matching her with an organ donor.
  • Pharmacists could use the database to contact providers and recommend newly available medications or alternative treatments for particular patients.

Aggregate data, which is data that has been stripped of individual identifiers, is also a valuable source for harvesting. In contrast to individually identifiable information, the privacy concerns are minimal, as the data is not connected with a particular patient.
The following are some examples of how aggregate data can be used:

  • Health care providers can study the data to analyze past usages of medical products and services and detect seasonal trends, enabling them to predict future needs and ensure that the proper level of supplies will be available.
  • Outcomes of groups of similarly afflicted patients treated with different drug regimens can be compared to determine which treatments statistically work better and are most cost effective. In addition, studies can be made to better predict the effects of diverse drug combinations and interactions.
  • Health care organizations and insurers can use aggregate data to determine if physicians are ordering unnecessary tests or treatments, by comparing the physician's patient profiles with the aggregate profiles of patients with similar conditions.
  • Health care providers can use the same data to contest a payer's refusal to cover charges for services deemed unnecessary, if the data shows that such tests or treatments are standard among patients with the same condition.

A Delicate Balance

Although there are numerous benefits tied to harvesting medical data, it does present some additional risks, especially to patient privacy. This risk varies, depending on the type of information used and analyses being performed. Notably, privacy concerns are more pronounced where individually identifiable information is involved.

Accordingly, it makes sense to place some legal restrictions on how patient data may be harvested. However, these restrictions should not be allowed to eclipse the many advantageous uses of patient databases. Legislators must recognize that a little loss of privacy, when weighed against the significant benefits of data harvesting, may prove to be very healthy for patients and the public in general.

For a more in-depth look at the changing legal landscape, data harvesting in practice and additional references, see Part 2 below.

1 Secretary of Health and Human Services, Confidentiality of Individually Identifiable Health Information, Recommendations of the Secretary submitted to Congress, September 11, 1997.

PART II (on-line only)

The Changing Legal Landscape for Data Harvesting

The recent enactment of the Health Insurance Portability and Accountability Act of 1996 ("HIPAA")1 promises to expand federal protection of patient data. Currently, only specialized classes of information, such as substance-abuse records or data held by federal agencies, are protected at the federal level. However, HIPAA seeks to establish national standards for electronically transmitting medical data. As part of those provisions, the Act charges the Health and Human Services Department (HHS) with making recommendations for improving confidentiality protections for individually identifiable health information.

The HHS report, released to Congress last year, notes that current state-level protections are inadequate, and argues for the establishment of federal privacy standards.2 If the recommendations are adopted, a legal duty of confidentiality would be imposed on those who provide and pay for health care, as well as on other organizations that receive patient data from them. Further, collected data could only be used for purposes compatible with and directly related to the purposes for which the information was collected.

The report also advocates that organizations holding health information be required to protect against wrongful disclosure of that information through the use of effective security measures. Included in the report are suggested civil and criminal sanctions for misuse of medical information.

The proposed standards also set out patient rights with respect to their records. Further, the HHS report includes a controversial recommendation for permitting limited disclosures of health information for certain "national priority activities," such as law enforcement activities.

Although a bill to implement the HHS recommendations has yet to be introduced, several other bills addressing health care privacy issues have been proposed. Pending bills on this issue include the "Fair Health Information Practices Act," H.R. 52 ("FHIPA"), and the "Medical Information Privacy and Security Act," S. 1368 ("MIPSA"), as well as a number of bills targeting more narrow groups of information, such as genetic data or information held by insurance companies. In addition, Senator Bennett (R-Utah) is expected to introduce a medical records bill early in the second session of Congress.

One significant difference between FHIPA and MIPSA is the effect a new federal law would have on state privacy laws. FHIPA would prevent states from imposing any additional requirements on "trustees" of health information (e.g., providers and payers), with the exception that state laws dealing with public or mental health would not be preempted. On the other hand, both MIPSA and the HHS report would provide a "floor" of protection, leaving individual states free to implement more stringent privacy requirements.

Until Congress sets a uniform federal standard, data privacy will continue to be governed by an inconsistent patchwork of state laws that vary considerably in strength and scope. In fact, only two states have enacted the so-called Uniform Health-Care Information Act since its adoption in 1985. That Act generally requires signed patient consent before a patient's medical information may be released.3

Some states protect only narrow categories of health information, such as records related to mental health or substance abuse treatment.4Other states, such as California5 and Florida,6 have enacted relatively comprehensive laws on the subject. Additional protection against the wrongful disclosure of confidential medical information may be available under judicially-created privacy doctrines.

Often these state laws are unclear in their applicability to data harvesting activities. Further, given the increasingly interstate nature of the health care field, the task of applying varying state laws is made even more difficult.

Perhaps the best solution for both patients and their health care providers is to establish a uniform federal standard that will balance the many beneficial uses of medical data and the interests of protecting patient confidentiality. In order to keep that balance, however, states should not be permitted to impose additional burdens on providers and payers (as recommended by the MIPSA bill and the HHS report). Congress should carefully consider the advantages of data access and recognize that a little loss of privacy may prove to be very healthy for patients and the public in general.

Data Harvesting in Practice

  • In a recent study of diabetic and "frail elderly" patients, pharmacists were able to identify 197 "drug-related" problems during a total of 136 patient visits. In 30 percent of the patients, pharmacists recommended discontinuation of at least one medication. In nearly all of those cases, the physician agreed to the change, resulting in estimated savings of $100 per patient per year, and presumably better outcomes for the patient. (Reuters, Jan. 15, 1997)
  • Sentara Health System studied its individually identifiable data and found that doctors were placing multiple orders for sputum cultures for the same patient in attempt to speed lab results. After redesigning the process to require test results to be delivered within two hours, Sentara ultimately reduced the mortality rate of pneumonia patients from 12 to 9 percent. The average hospital stay also dropped to one week, generating savings of $2,000 per case. (CIO, June 15, 1997)
  • An on-line asthma management program of the University of Pennsylvania Health System links primary care physicians, pharmacists, case managers, and other health care providers. By tracking compliance of over 30,000 asthma patients, the program reduces drug misuse, duplicate therapies, excessive dosing, and potential drug interactions. (Reuters, July 16, 1997)
  1. Pub. L. No. 104-191 (August 21, 1996).
  2. Secretary of Health and Human Services, Confidentiality of Individually Identifiable Health Information, Recommendations of the Secretary submitted to Congress, Sept. 11, 1997 (hereinafter "HHS Report").
  3. 9 Part I, U.L.A. 475 (1988), adopted by Montana and Washington.
  4. See, e.g., 740 ILCS 110/3 (West 1993).
  5. Cal. Civ. Code 56 (1997).
  6. 6 Fla. Stat. Ann 395.3025(4) (1996).

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access