Feasibility of Including Physiological Parameters in Medicare Claims Data for the Treatment of Chronic Conditions

An Opportunity to Improve Quality of Care in the Medicare Program Using Enhanced Administrative Data

See also :Comments to Medicare Modernization Act of 2003, Section 1013 Symposium, May 21, 2004

MTPPI is requesting comments on whether implementation of Medicare Prescription Drug, Improvement, and Modernization Act (MMA) should include a provision that relevant physiological information be included in future Medicare claims data for chronic conditions for visits where a prescription is administered or where a prescription for a drug is written. Enhancement of such data could be accomplished through the imposition of such a requirement using physician outpatient claims for these services or by merging these claims with laboratory service values. Including a physiological parameter might be especially useful to examine "off label" use or in situations where there is clinical controversy and where documenting a physiologic endpoint is needed.

We would be interested in receiving your comments on the following two related questions:

1) Would such a requirement enhance the informational content of the health care system and its treatment outcomes of Medicare beneficiaries?

2) What are the possible limitations or obstacles to such a strategy?

As an example of the usefulness of such information, we are studying the claims data for epoetin therapy which include hematocrit values contained in Medicare administrative databases for purposes of epoetin billing. This information has been shown to enhance understanding of the therapy, outcomes, and cost-effectiveness associated with an expensive drug. However, unlike controlled clinical trials, our observational analysis of administrative data must grapple with the confounding effects of unobserved events. If done well, analysis of such information can provide important insight into the 'real-world' risks and benefits of new interventions.

Aside from anemia related dialysis and chemotherapy, such a requirement may be applicable to certain chronic or long-term conditions (e.g., arthritis, asthma, congestive heart failure, depression, diabetes, etc.). When a prescription is written the provider makes a judgment based on whether the patient's condition requires continuation of therapy and whether dose adjustments are warranted for the next cycle of therapy. This decision is based on a physical examination and/or physiologic measurement(s) (e.g. laboratory value(s), a measurement that served as either an endpoint or surrogate endpoint when the product was approved by FDA). The provider submits a claim for this service, so the extra burden should be minimal.

Summarized in the table below are comments on the feasibility of including physiological parameters in large administrative databases received by MTPPI from numerous experts in the private and public sectors. Please provide us with your name and affiliation and any relevant comments you may have with regard to this issue. We thank you in advance for your contribution to this project.

Table 1. Inclusion of Physiological Parameters in Medicare Prescription Drug, Improvement, and Modernization Act (MMA)

Potential Advantages

Potential Disadvantages

MMA Specific Issues

1. Improved understanding of the effectiveness and efficiency of different treatments

– determine the relationship between intervention and outcome

1. Logistical difficulties:

– few medical practices have electronic medical records

– physicians do not have unique provider ID codes

– many physician practices do not conduct their own billing. ^d

1. New Medicare Rx benefit program has provisions for demonstration projects and changes in electronic information sharing by 2008, thus offering a mechanism to support inclusion of such information.

2.Pharmacoepidemiologic applications

– adverse events for all exposed patients

2. Political considerations:

– professional societies have repeatedly lobbied against any regulatory efforts to improve computer records and billing in physician practices

– trade groups might also object

2. Mandate inclusion of physiological parameters?

– PBM could make this decision

– Could be one of the conditions for Medicare to establish coverage

– mandating might cause significant administrative problems

3. Improved quality of care

– facilitate more rational treatments

– promote use of evidence-based guidelines

– assess extent to which physicians are following validated guidelines

–counterdetailing; information provided to all prescribers

3. Cost and burden:

– time and labor intensive

– cost cannot be borne by physician practices

– PBMs and stand alone drug plans do not now collect or have the capacity to collect physiological information; they would probably find it to be an onerous undertaking for which they

would want to be paid.

– geriatricians, who care for large numbers of the elderly with chronic conditions, are most overwhelmed and most likely to be burdened by additional reporting requirements

– samples or specimen may be obtained at the visit to the physician, but the tests may not be completed for some time after the end of the visit^e

– increased costs in terms of IT requirements and data tracking

3. First we must require Medicare providers to include a relevant diagnosis with prescriptions, as is the case in nursing homes.

4. Generate hypotheses to test in a well- designed clinical trial^a

4. Patient privacy issues

– HIPAA implications of sharing confidential information among various service providers.

– concerns about who and how the data would be used (e.g., role of cost-effectiveness studies by the same agency setting payment rates)

4. Test results generally are not submitted with the claim. ^g

5. Identify all indications and uses for prescribed drugs

5. What information would be collected?

– Would it change depending on the drug (i.e., HepC geno-testing for any HepC treatment, hematocrit for epoetin, culture and sensitivity testing for antibiotics ^f

– How large would the prescription form need to be to accommodate this information?

– What happens if the physician doesn't have lab results yet at the time that the prescription is being written?

6. Logistical considerations: to the extent results are already transmitted electronically to the provider, the added costs of transmitting the information to Medicare would not be great.

7. Derive clinical decision tools

The opinions expressed in this document do not necessarily reflect the position of MTPPI .

NOTES:

a Such a hypothesis and the resultant study design would be more pertinent being guided by the real-world real-time observational data. Caveat is that it depends on on the amount of confounding and whether the confounders can be controlled for with the level of data that you hope to collect.

b These demonstrations are exceptions to the rule and focus on the cost/benefit of various initiatives for a specific condition (diabetes, CHF, etc).

c Use of epoetin is an excellent case study and represents much of the practice of medicine since epoetin therapy is aimed at improving a biological endpoint (in this case, hematocrit). It is rational to know the starting and final hematocrit based on treatment to predict the likely health outcome.

d Physician medical practices often hire outside services to conduct the electronic billing on their behalf.

e Would the physician be required to wait for the results before making a claim for payment? Some tests take some time for the results to be known. How long would physicians be expected to wait before submitting a claim? Would they tolerate a wait of more than a few days?

f For example, for some such as SSRIs and cancer drugs, there are no definite quantitative data.

g Traditionally, with drugs administered in a hospital our outpatient setting, Medicare coverage for laboratory tests can be provided as indicated by the drug label or professional standards of practice. For example, HER-2 testing is covered for determining which patients will receive Herceptin therapy, and then cardiac monitoring tests are covered to detect any early signs of cardiotoxicity. However, the results of the tests generally are not submitted with the claim (only the information that a test was taken) so they could not be monitored. Because the Medicare program is administered at the local level, coverage for drug-related ancillary services can differ by local carrier or contractor.

Potential Advantages	Potential Disadvantages	MMA Specific Issues
1. Improved understanding of the effectiveness and efficiency of different treatments – determine the relationship between intervention and outcome	1. Logistical difficulties: – few medical practices have electronic medical records – physicians do not have unique provider ID codes – many physician practices do not conduct their own billing. ^d	1. New Medicare Rx benefit program has provisions for demonstration projects and changes in electronic information sharing by 2008, thus offering a mechanism to support inclusion of such information.
2.Pharmacoepidemiologic applications – adverse events for all exposed patients	2. Political considerations: – professional societies have repeatedly lobbied against any regulatory efforts to improve computer records and billing in physician practices – trade groups might also object	2. Mandate inclusion of physiological parameters? – PBM could make this decision – Could be one of the conditions for Medicare to establish coverage – mandating might cause significant administrative problems
3. Improved quality of care – facilitate more rational treatments – promote use of evidence-based guidelines – assess extent to which physicians are following validated guidelines –counterdetailing; information provided to all prescribers	3. Cost and burden: – time and labor intensive – cost cannot be borne by physician practices – PBMs and stand alone drug plans do not now collect or have the capacity to collect physiological information; they would probably find it to be an onerous undertaking for which they would want to be paid. – geriatricians, who care for large numbers of the elderly with chronic conditions, are most overwhelmed and most likely to be burdened by additional reporting requirements – samples or specimen may be obtained at the visit to the physician, but the tests may not be completed for some time after the end of the visit^e – increased costs in terms of IT requirements and data tracking	3. First we must require Medicare providers to include a relevant diagnosis with prescriptions, as is the case in nursing homes.
4. Generate hypotheses to test in a well- designed clinical trial^a	4. Patient privacy issues – HIPAA implications of sharing confidential information among various service providers. – concerns about who and how the data would be used (e.g., role of cost-effectiveness studies by the same agency setting payment rates)	4. Test results generally are not submitted with the claim. ^g
5. Identify all indications and uses for prescribed drugs	5. What information would be collected? – Would it change depending on the drug (i.e., HepC geno-testing for any HepC treatment, hematocrit for epoetin, culture and sensitivity testing for antibiotics ^f – How large would the prescription form need to be to accommodate this information? – What happens if the physician doesn't have lab results yet at the time that the prescription is being written?
6. Logistical considerations: to the extent results are already transmitted electronically to the provider, the added costs of transmitting the information to Medicare would not be great.
7. Derive clinical decision tools

An Opportunity to Improve Quality of Care in the Medicare Program Using Enhanced Administrative Data