Microsoft word - prior authorization testimony follow up.docx

Thank you for al owing me the opportunity to testify on 4/4/2013 before the House Committee on Health Care regarding prior authorizations. This message is in response to the Committee’s request for me to fol ow up with a summary of suggestions for improvements and to provide additional information Improvement Opportunities
Multiple year authorizations
Although many chronic medical conditions can be treated with first line alternatives that do not require
authorization review, the prior authorization process al ows approval for second or third line alternatives for appropriate cases. Industry standard practice has been to approve such authorizations one year at a time. However, a BCBSVT analysis revealed that these cases typical y continue to meet approval criteria ongoing in more than 97% of reviews. We became the first health plan in the country to have our Pharmacy Benefit Management system issue such authorizations for three year intervals.
This is projected to remove more than 17,000 prior authorizations from the workload of Vermont prescribers without significantly increasing cost of care. At the same time, this approach continues to guide the majority of cases to the most cost effective alternatives and preserves incremental pricing discounts available for selected medications when prior authorization requirements are in place. We
encourage Medicaid and other health plans to consider adopting this approach. We do not believe
this requires legislative action.
“Instant” authorizations and online authorization systems
BCBSVT has posted medication-specific authorization request forms that indicate exactly what
information is required for review. These forms usual y make it unnecessary to submit copies of clinical records, eliminate repeated requests for additional information, and al ow rapid approval when criteria are met. Once familiar with the forms, physician offices find this approach reduces administrative burden considerably compared to sending copies of clinical records. This is similar to the experience in North Carolina in which use of preferred proton pump inhibitor medications quadrupled with implementation of an “instant authorization” form that only requested key information. Wegner SE,Trygstad TK, Dobson LA Jr, Lawrence WWJr, Steiner BD. A physician-friendly alternative to prior authorizationfor prescription drugs. Am J Managed Care. 2009 Dec 1; 15(12):e115-22.
BCBSVT is implementing an online system for medical-surgical authorizations that is targeted to be available beginning in April 2013. Interactive data entry wil al ow automated approval for requests meeting specified criteria. We plan to increase automated approval capabilities over time. This system wil also al ow medical offices to confirm the status of current requests online, which will reduce office administrative burden associated with duplicate submissions and inquiry telephone cal s. Online authorizations tailored to the needs of specific requests have already been implemented for our pharmacy authorizations and for our high technology imaging authorizations. However, use of online systems varies across provider offices. There may be opportunities to embed access to such online systems into the decision support functionality of electronic health records. We suggest that Act 171
from last session be revised to clarify that use of electronic authorization systems designed to elicit
specific information is not prohibited (see note at end of this message).
Smart authorizations
Effective 7/1/2013, BCBSVT wil activate a system that uses medical-surgical claims information to
automate approval for a variety of medication authorizations in which the necessary clinical data is inherent in the existing claims. We wil require a number of months with this system operational before we can report on the magnitude of administrative burden reduction achieved. We encourage Medicaid
and other health plans to consider adopting similar automation. We do not believe this requires
legislative action.
Integration of medication criteria with electronic order entry systems
Considerable rework occurs when a patient arrives at a pharmacy to fil a prescription only to have that
prevented because the prescriber did not obtain prior authorization at the time of prescribing. Moving the criteria and informed decision factors into the care process at the time of the initial prescribing would reduce administrative burden and administrative cost. BCBSVT is interested in working with practices to achieve efficiencies integrating authorization criteria into decision support aids in electronic prescription system and/or electronic medical record systems. A variety of contractual agreements could be considered. Although the office practice administrative savings may pay for implementation and generate savings for the physicians, practices are justifiably wary of doing this in absence of documented success in other practices. In addition, currently instal ed electronic systems have not been wel designed to support such decision support and may require expensive programming to modify.
Adding this functionality would be useful for providers who may be required to accept and manage financial and clinical accountability as payment reforms such as ACO population-based payments are implemented. We encourage the Green Mountain Care Board to use grant funding to support pilot
implementation of integrating criteria into electronic order entry systems in a representative sample
of practices.
Provider office administrative efficiencies
We note that a recent publication documented annual per physician cost of prior authorization
requirements to range from $2,161 to $3,430. Morley CP, Badolato DJ, Hickner J, Epling JW. The impact ofprior authorization requirements on primary care physicians' offices: report of two parallel network studies. J AmBoard Fam Med. 2013 Jan-Feb;26(1):93-5. While respectful of the burdens faced by practices and empathetic with the frustrations that often develop, we also find that, somewhat unexpectedly, a number of practitioners make this much more burdensome than necessary. In some situations, simple but effective steps are available that can work wel for the self interest of the practice while better meeting the needs of their patients.
 Prescribe medications more often that do not require prior authorization. For BCBSVT, 98.1% of prescription claims are fil ed with no need for prior authorization (data below). Even paper listscan help achieve a large percentage of this goal. Use of Epocrates software (available for free)can identify authorization requirements specific to many health plans. BCBSVT pays to haveformulary information available on Epocrates.
 When the patient requires a medication that does need prior authorization, complete the process during the office visit rather than waiting for a pharmacy to send notification severaldays later. Again, even paper based lists can accomplish a substantial portion of this. Integrationof authorization information into electronic prescribing systems could be even more effective,as discussed above.
 When requesting an authorization, send the pertinent information the first time rather than waiting until a denial determination forces the issue. In 2012, BCBSVT identified multiple officesthat consistently generated unnecessary work for themselves by not sending clinical informationwith authorization requests. Educational outreach to these offices substantial y reduced thewasted effort and reduced administrative burden for these offices, but has not yet eliminatedthe problem.
 It may be useful to convene a workgroup of providers who can share best practice approaches to reducing self-imposed unnecessary work. This can help even while longer term improvementsare being designed and implemented. Sometimes simple initiatives can require resources toimplement but can pay for themselves repeatedly through subsequent reductions in workload.
New contractual relationships between health plans and providers
The Green Mountain Care Board has announced intent to advance population-based payment reform
using Accountable Care Organization (ACO) contracts. When ACOs assume financial and clinical “risk”, they implement innovative systems to improve cost-effective care delivery and improve outcomes. In systems such as Intermountain Health Care in Utah, clinical guidelines and authorization criteria have moved out of the health plan realm (for targeted internal providers but not for out-of-area non-contracted providers) and have been integrated into the clinical delivery system as welcome tools rather than as imposed burdens. We suggest that the Legislature may not need to take immediate action
now. It may be wise to allow the ACO contracts and systems to develop effective mechanisms to
improve outcomes while constraining costs.
Transmission of current medication authorizations to next health plan during transitions
When large employer groups change from one health insurance plan, information held by the former
health plan regarding approved authorizations for services and prescriptions can be transmitted to the new health plan. Typical y this is accomplished using an electronic file. When that data interchange occurs, the new health plan has the opportunity to continue the authorizations, eliminating the need for prescribers to submit new authorization requests and new copies of clinical information. When that data interchange does not occur, re-work is required and gaps in care may result. The legislature could
consider convening a work group to evaluate the costs and benefits of requiring health plans to share
authorization information in such circumstances.
Denial Rates
Sentinel effect to consider when evaluating “denial” rates
Although “denial” rates command attention, a denial rate does not serve as a reasonable index for the
effect of a prior authorization requirement. After implementation of an authorization requirement, providers learn which alternatives do not require authorization, and they learn what factors lead to approval decisions. This learning, known as the sentinel effect, results in shifts in treatment decisions
separately from the changes induced by direct review interventions. Requests for treatments needing review decrease. An effective and efficient prior authorization system could have very few denials and a low number of reviews. Removal of an authorization requirement when very low denials are generated frequently results in a disproportionate increase in utilization and cost. As I testified, that happened in Vermont when BCBSVT removed PA requirements for 22 medications that had low denial rates. Within the next year, utilization approximately doubled for these medications and cost increased more than $800,000 per year. Hence, evaluation of the effect of a prior authorization requirement should use
indicators other than denial rates.
One reasonable approach is to consider shifts in market share of the various alternatives after implementation of an authorization requirement. For example, imposition of a prior authorization requirement for cefuroxime (an antibiotic) resulted in a reduction of prescriptions from 5,538 in three months to 1,036 in three months. This represents an 81% reduction in prescriptions, with the 8.5% rejection/redirection rate affecting only the reduced number of prescriptions available for authorization review. There was a corresponding increase in appropriate alternative antibiotic prescriptions. Removal of the requirement was fol owed by an increase to 3,961 cefuroxime prescriptions in three months. That represents a 3.8 times increase, consistent with other reported sentinel effect situations. Kahn HR,Chinitz DP, Waitman DA, Kahan E. When gatekeepers meet the sentinel: the impact of a prior authorizationrequirement for cefuroxime on the prescribing behaviour of community-based physicians. Br J Clin Pharmacol.
2006 Mar; 61(3):341-4.
Another reasonable approach is to compare utilization rates and costs in a health plan that uses an authorization requirement to concurrent rates in another comparable plan that does not use the authorization requirement. Our imaging program uses this methodology, see below.
BCBSVT Prescription Drugs
2012 calendar year:
17,570 authorization requests (1.1% of claims) 2,815 denied (16% of reviews – multiple reasons including eligibility, safety criteria, and medical criteria; this is 0.18% of medication claims) Many of the denied requests resulted in fil of a different prescription in the same class. When that happens, the cost impact (decreased or increased cost) are calculated as the difference between the initial y requested prescription and the fil ed prescription. We have requested specific 2012 data on the redirection rate. However, that figure is not available for this response.
The direct impact of denial/redirection was $2.85M in 2012. With a sentinel effect of 2x to 3x, that projects to an effect of $5.7M to $8.5M annually. Primary care providers issued 40% of the
prescriptions and accounted for 51% of the dollar impact of the authorization program.
It is important to note that medication prior authorizations may be triggered by patient safety criteria.
The electronic pharmacy payment system may block fil of a prescription when cumulative dose of ingredients would be expected to be toxic. For example, multiple medications containing acetaminophen may cause liver failure if taken together. This is particularly dangerous when prescriptions are issued by multiple prescribers and fil ed at multiple pharmacies. A prior authorization review can al ow appropriate exceptions while assuring that toxicities are properly evaluated.
In addition, it is important to notice that the sentinel effect of the authorization program shifts many prescriptions to cost effective alternatives that do not require prior authorization and that often have lower copayment requirements for the patient. This financial benefit to members who end up with
lower out of pocket cost (and with lower premium payments over time) should also be considered when evaluating the impact of prior authorization programs.
BCBSVT High Tech Imaging Authorizations
In 2012
95.8% approved (rate varies considerably by practice and physician) 2.6% redirected (62% of directly impacted requests)1.6% denied (38% of directly impacted requests) Due to the sentinel effect, the rate of denials has progressively decreased since initiation of the program. Using calculations based on comparison to utilization rates in comparable health plans without the program, the net effect of the program (including requests redirected to more expensive studies) comes to $6.6M/year. This is consistent with calculations based on direct intervention impact and
It is important to notice that this imaging authorization program also addresses patient safety.
Estimated cumulative dose of radiation is calculated for each patient across time and across facilities.
Based on 2008-2010 data in our Vermont data, approximately 560 exams annually receive safety
intervention because of the high cumulative radiation exposure.
BCCBSVT denial rates: administrative compared to member impact
In data filing for 2012 required by Act 150 and recently posted to the DFR web site, denial volume is
 Administrative denials: denials with no member impact, duplicate claim check, invalid place of service, invalid coding, refil too soon, member not active, member hold harmless, otheradministrative denials. 94% of denials.
 Member impact denials: not covered, excluded, benefit limits met, paid at lower level of benefit, prior approval was denied, not FDA approved, step and quantity limits, out of network,investigational/experimental, waiting periods, not medical y necessary, other member impactdenials. 6% of denials.
Interventions to reduce administrative burdens in practices may not have the desired result if focused (as the proposed bil is) only in authorization denials based on investigational, experimental, and medical
necessity criteria (subset of Member Impact denials). We suggest that S.40/H.77 not be approved.
Convening a group of stakeholders to evaluate clinically and financially responsible efficiencies and
practical alternatives to reduce administrative burden may be useful.
Note regarding opportunity to revise ACT 171
In the last legislative session, Act 171 amended Sec. 11h. 18 V.S.A. § 9418b to require development of
uniform prior authorization forms. This attempt at simplification runs the risk of inadvertently causing increased administrative burden for physician offices due to the need to supplement such forms with copies of clinical information. That would be a step backward from the advances documented above. It also may impede implementation of medication authorization mechanisms integrated at point of care within electronic order entry systems. While BCBSVT plans to accept uniform request forms as required, we suggest that the Legislature consider revising Act 171 to clarify that nothing in the Act prohibits
use of more specific prior authorization forms or electronic authorization systems as long as the
alternative of using a uniform request form is also offered.
P.O. Box 186 • Montpelier, VT 05601-0186

Source: http://www2.leg.state.vt.us/CommitteeDocs/House%20Health%20Care/Bills/H.77/4-17-2013~Robert%20Wheeler~H.77~Dr%20Wheeler%20BCBS%20prior%20authorization%20testimony%20follow%20up.pdf