© Copyright 2013 Willis Limited / Willis Re Inc. All rights reserved: No part of this publication may be reproduced, disseminated, distributed, stored in a retrieval system, transmitted or otherwise transferred in any form or by any means, whether electronic, mechanical, photocopying, recording, or otherwise, without the permission of Willis Limited / Willis Re Inc. Some information contained in this document may be compiled from third party sources and we do not guarantee and are not responsible for the accuracy of such. This document is for general guidance only and is not intended to be relied upon. Any action based on or in connection with anything contained herein should be taken only after obtaining specific advice from independent professional advisors of your choice. The views expressed in this document are not necessarily those of Willis Limited / Willis Re Inc., its parent companies, sister companies, subsidiaries or affiliates (hereinafter “Willis”). Willis is not responsible for the accuracy or completeness of the contents herein and expressly disclaims any responsibility or liability for the reader's application of any of the contents herein to any analysis or other matter, or for any results or conclusions based upon, arising from or in connection with the contents herein, nor do the contents herein guarantee, and should not be construed to guarantee, any particular result or outcome. Willis accepts no responsibility for the content or quality of any third party websites to which we refer.

In this issue From the editors

Catastrophe modeling – lifting the lid on Workers’ Compensation models and how data quality affects outputs

2 In a continuation of our efforts to deliver cutting edge insight into reinsurance issues facing Workers’ Compensation writers, this edition of the Willis Re Workers’ Compensation Review takes a detailed look at catastrophe modeling. Catastrophe modeling is a much more mature science in the Property world than in ours, as data quality has advanced significantly over the last decade. Our lead article focuses on how the different levels of data that Workers’ Compensation carriers collect from insureds could ultimately result in potential cost savings on reinsurance programs. For those who were unable to attend the recent NCCI AIS conference or did not have time to wade through the various presentations, we provide a quick synopsis of the key takeaways from Dennis Mealy’s final State of the Line presentation. Medical inflation is one of the most talked about subjects in our industry and for reinsurers it is the most important topic. Per person reinsurance generally attaches somewhere around $1 million. Over 85% of claim costs for claims above $1 million arise from medical costs, so it’s easy to see why medical inflation is such a high profile issue in the reinsurance world. In this edition, we consider what the future may hold and project some potentially positive trends for the industry. We also include two guest articles this month, one looking at the issue of mapping procedure codes from ICD-9 to ICD-10 and the potential effect on Workers’ Compensation carriers, and the other taking a look at physician drug dispensing and how the costs escalate through the supply chain with some quite remarkable implications. We thank consultants Point B, Edifecs and PMSI for their contributions to this edition.

Conference highlights: National Council on Compensation Insurance (NCCI) May 16, 2013, Orlando, Florida

5

Medical disinflation – theory or reality? 6

Guest Article: Point B, Inc. and Edifecs – Are you prepared for the Workers’ Compensation “Drag Along” provision in U.S. Healthcare Reform?

8

Guest Article: PMSI – The growth and management of physician dispensing

10

In our next issue… With the large 1/1 renewal season looming, we provide analysis of the reinsurance industry that will help let our clients know what to expect during this year’s renewals.

July 2013

WORKERS’ COMPENSATION

REVIEW

2

Catastrophe modeling – lifting the lid on Workers’ Compensation models and how data quality affects outputs

How does a model calculate loss? When your reinsurance intermediary presents you with your yearly 1-in-100 and 1-in-250 modeled losses, have you ever questioned how these amounts are derived or what you can do to reduce the uncertainty in the outputs? Simply put, catastrophe models all begin with damage to buildings (the property module). Workers’ Compensation models for both earthquake and terrorism are built upon the same event, hazard and vulnerability modules as Property modeling. The severity of the building damage is correlated with the number of “lives” affected within each building (the human module). These lives are then developed into cost severities by allocating them into different injury types before assigning each type an average value (the financial module). Aggregate these items together and you get a deterministic loss estimate. Throw thousands of loss simulations at the same portfolio and you obtain a probabilistic loss for a 100 year or 250 year event. This number then changes based on the time of day, which is why most companies look at both their peak exposures as well as their temporal exposures (an average of both peak and off peak). The injury severity amounts, which vary by state, are derived from various industry data sources including, but not limited to, the National Council on Compensation Insurance (NCCI) Annual Statistical Bulletin and the Hospital Cost Utilization Project (HCUP).

What data makes a difference in modeling outputs? The human vulnerability and financial modules add another layer of complexity and uncertainty to the already untested and uncertain earthquake and terrorism property models. Large uncertainties in modeling results can initially overwhelm and reduce confidence in risk managers and (re)insurance executives, but that doesn’t mean significant value can’t be derived from the modeling output. Focusing on accurate exposure information ranging from employee demographics, employee locations and building characteristics can reduce uncertainty and provide companies a better understanding of their own book.

Key components to the Workers’ Compensation model Employee information (payroll vs. employee count)

• If you only collect payroll information, the model will determine an employee count based on the occupation code and geography. – For example, if on average, employees in California have $50,000 in payroll, but a client portfolio has an average

payroll of $60,000, the model will automatically overestimate the employee count by 20%. • Even when employee counts are reported, they are not always representative of the exposed risk. Often, part time

employees will be represented the same as full-time employees in the underlying data. If the information is collected accurately and consistently, one can make adjustments for part-time employees, such as 1 part time = .5 full time.

• In addition to part-time employees, businesses with work shifts can be inaccurately represented. The maximum number of employees working at any one time (MAOT), unless otherwise defined, is determined by the occupation of the employees. These maximums typically range from 65% to 90%. Any divergence from these assumed values can greatly impact the employees subjected to modeled events.

Employee locations

• Collecting complete and accurate location information for employees is more significantly important than the underlying hazard variance for certain perils from location to location. – Earthquake ground shaking intensity can vary from one end of a street to the other end. Soil type and the

corresponding soil amplification of the ground shaking can also vary significantly within a zip code / street.

3

• Having street level or latitude / longitude information for all employees is ideal. – Some rating agencies, such as A.M. Best, have instituted

geocoding surcharges (up to a 25% penalty on the Terrorism PML BCAR charge calculation) for portfolios having less than 90% street address or geocoding.

– Catastrophe models assume inventory of various primary construction characteristics for a specific building occupancy when the primary characteristics of input risk are “unknown.” These assumptions will vary based on location of the risk (e.g., rural, urban or city center).

• Employee concentrations – Public employment agencies, temporary staffing

agencies, company headquarters, colleges / universities / hospitals, etc. • Large numbers of employees can be impacted by

individual events when reported in aggregate (rather than at separate individual locations).

• If these concentrations happen to be located in high risk areas, the modelled impact is magnified.

• Special considerations – Mobile employees are not in a fixed location, so their

loss potential cannot be accurately estimated from static data.

– Employees who work from home may be reported as though they were at the office location, which may lead to overestimation of loss.

– Employees not located in buildings (such as farmers) are not accurately modelled, since the commercial models estimate losses based upon building damageability.

Primary building characteristics

• Historically, Workers’ Compensation companies have not included building characteristics in their modeling data. Companies providing Workers’ Compensation coverage often do not insure the property so obtaining this information can be challenging. However, today more and more companies cross reference their multi-line policy data or ask their agents to provide building characteristic information when underwriting polices.

• Significant damage to buildings, especially collapse, is the primary cause of employee injury. – Construction, number of stories, year built and occupancy are the primary characteristics used to determine the

vulnerability of employees during an event. – In our opinion, construction is the most significant component in modelling.

• Masonry and concrete buildings are more susceptible to collapse and damage while wood and steel buildings are less susceptible.

– Number of stories is the second most significant component. • Building height is used in conjunction with construction to determine shaking intensity levels specific to a

building during earthquake event; typically, shorter buildings with one to three stories will perform better than taller (more slender) buildings when the ground shakes.

– Year built • Year built is included in the model to reflect changing building code requirements and construction practices

over time. – Occupancy

• In addition to determining employee characteristics, building occupancy can be used to estimate the building inventory when other primary characteristics are “unknown.”

4

How do the primary building characteristics impact the modeled losses? Consider a hypothetical portfolio of California risks evenly distributed across the state, where all of the primary characteristics are “unknown.” On a state-wide basis, the Workers’ Compensation Average Annual Loss (AAL) to this portfolio would be $3,289,037 (see table below). By adding in a construction type of Steel Frame for all buildings, the AAL is reduced by 89% to $362,435. By adding in a low-rise height to the Steel Frame building, the AAL decreases further to $349,791. And finally, adding in a newer date of construction to the low-rise Steel Frame building reduces the portfolio’s AAL $320,089.

On the contrary, the addition of known primary characteristics can also increase the portfolio’s AAL. By adding in a construction type of Ductile Reinforced Concrete Frame for all buildings, the AAL is increased by 101% (to $6,621,787). Adding in a high-rise height to the Ductile Reinforced Concrete Frame building raises the AAL to $11,988,031. Lastly, adding in an older date of construction to the high-rise Ductile Reinforced Concrete Frame building increases the same portfolio’s AAL to $12,017,989. Considering this synthetic portfolio, one can see that modeled losses can vary significantly depending on the primary building characteristics provided in the underlying data.

So what does this mean as a reinsurance buyer? In the Property world, reinsurance pricing is a function of the mean loss plus a percentage of standard deviation. More accurate data reduces model uncertainty, lowers the standard deviation and, in turn, lowers reinsurance pricing because reinsurers accept a lower margin in exchange for less uncertainty. While Workers’ Compensation pricing is still predominantly a capacity priced product, as models become more sophisticated and reinsurers use technical pricing, companies with better data quality will in turn feel the financial benefit. Increased data quality is also appealing to the huge influx of capital coming from non-traditional reinsurers. These reinsurers offer fully collateralized limits for their clients, who benefit from a significant security advantage for long-tail lines such as Workers’ Compensation. Aside from the reinsurance purchase, rating agencies also examine data quality when determining a company’s rating. Improved data quality should also help risk managers better manage their exposures and loss potential on a daily basis. While Workers’ Compensation modeling is still in its infancy and largely untested, the benefits of better data quality impact all industry stakeholders. Please contact your Willis Re Client Advocate® for how our analytics tools and services can help you understand your data better.

Average Annual Loss

% of "Unknown"

Average Annual Loss

% of "Unknown"

Average Annual Loss

% of "Unknown"

"Unknown" Primary Characteristics 3,289,037.79 100.00% 6,559.98 100.00% 1,593.60 100.00%Moment Steel Frame,

rest "unknown" 362,434.61 11.02% 262.60 4.00% 190.70 11.97%

Moment Steel Frame, low rise, rest "unknown" 349,790.54 10.64% 212.36 3.24% 173.85 10.91%

Moment Steel Frame, low rise, 2005 YR built 320,088.61 9.73% 192.99 2.94% 158.02 9.92%

Average Annual Loss

% of "Unknown"

Average Annual Loss

% of "Unknown"

Average Annual Loss

% of "Unknown"

"Unknown" Primary Characteristics 3,289,037.79 100.00% 6,559.98 100.00% 1,593.60 100.00%Ductile Reinforced Concrete Frame, rest

"unknown" 6,621,786.71 201.33% 4,969.76 75.76% 3,426.45 215.01%

Ductile Reinforced Concrete Frame, 20+ stories, rest "unknown" 11,988,031.32 364.48% 6,482.76 98.82% 5,055.97 317.27%

Ductile Reinforced Concrete Frame, 20+ stories, 1950 YR built 12,017,988.80 365.40% 6,580.63 100.31% 5,112.01 320.78%

State-wide Los Angeles – 90013 San Francisco – 94115

Synthetic California Workers' Compensation Portfolio

State-wide Los Angeles – 90013 San Francisco – 94115

5

Conference highlights: NCCI

National Council on Compensation Insurance (NCCI) held its Annual Issues Symposium (AIS) on May 16-17 in Orlando, Florida. The AIS is always well attended by clients, reinsurers, brokers and rating agencies, as the NCCI’s findings are the basis for most insurance pricing and bureau loss costs. The State of the Line Presentation outlines the NCCI’s analysis and conclusions from the past year and is unveiled at the AIS each year. The entire State of the Line presentation can be found here: https://www.ncci.com/nccimain/Events/MinutesPresentationsMaterials/Pages/NewsFromAIS2013.aspx Below, we briefly summarize the key highlights from the symposium: Market is hardening but still unprofitable CY Premium increased 9% in 2012 to $40B

• +3.8% growth in payroll (economic & wage increases)

• +1.1% change in bureau loss cost • +2.6% change in carrier discounting • +1.6% audits and other

CY combined ratio improved six points to 109% • 2012 CY ultimate net loss & LAE ratio of 81% • 2012 AY combined ratio of 108% • Combined ratio improvement still not enough to

be profitable due to the low investment climate (although this would have been deemed acceptable 10-15 years ago)

• Reserve deficiency increase (2% of reserves) implies real combined ratio is higher than 109% (2011 reserve deficiency grew by 1% of reserves)

Reserve deficiencies increased to $13B (11% of carried reserves)

• Tabular discount is $5.2B (reserve discount permitted on the indemnity portion of long term Workers’ Compensation claims)

• Should result in continued upward rate pressure • Reserve releases have almost all dried up, so the

unprofitable nature of the line is clearly showing Residual market grew by 50% (~$400M to ~$600M)

• Symptom of hardening market

Beneficial loss cost trends Frequency decline back in line with historical trend: -5% Severity growth moderates

• Indemnity +1% • Medical +3% • Growth in medical severity slowing due to physician fee

schedules and prescription management • Lower than medical CPI growth (4%)

Notes on the reinsurance market

• Reinsurers use higher medical trend than NCCI, which is a good argument to put to reinsurers when placing per person exposed layers

• Most reinsurers don’t use the full NCCI loss development tail which offsets the higher medical trend

NCCI general conclusions Market concerns

• Pace of economic recovery • Underwriting results • Low interest rates • Potential expansion of alternative systems (i.e., OK

adoption of non-subscription) • Healthcare reform

NCCI market positives • Premium increase • Return of frequency decline • Moderation of severity growth • Industry capital base

NOTE: Rate increases and combined ratios differ state by state; this is a nationwide average. These conclusions do not hold in every case.

6

Medical disinflation – theory or reality? The medical services industry is in the beginning stages of some major structural changes, the effect of which we anticipate will be to generate long term reductions in healthcare costs throughout the U.S. Healthcare industry. The Workers’ Compensation industry will share in these reductions. While we are not ignoring the significant long term inflationary pressures on the industry through medical costs (created by aging baby boomers, doctor shortages and other supply and demand issues) nor predicting medical deflation (as we are already seeing medical disinflation i.e., lower inflation rates), we fully expect medical inflation to continue but at lower rates than previously thought. According to the National Council on Compensation Insurance (NCCI), the average medical cost per claim for 2010 - 2012 increased by only 2.7% (2012 is preliminary). This is relative to a 5.7% increase for 2007 – 2009 and an average increase of 9.9% for 1997 – 2001. While a degree of causality has been attributed to the recent recession with which this trend of lower medical inflation coincides, there are other factors contributing to this phenomenon.

In order to look at what is causing this trend, we must first take a step back to look at the problem. Medicine, first and foremost, is about caring for people. However, it is also a business and a major part of the American economy. The medical system has historically operated under a business model which has resulted in medical costs being incredibly high relative to those seen internationally. Many believe the primary weakness of the business model is that health providers are not paid to cure people but rather to perform services and the more services provided, the more they are paid. As an industry, there is very little incentive for efficiency or to invest in low-cost technology. The cost can simply be passed on to the consumer. However, we are starting to see changes in the fundamental business model and we believe we are on the verge of a significant structural transformation, one which will benefit the industry's cost structure. This transformation is being driven by the pain felt as a result of the high medical inflation experienced for many years. Businesses are seeing their margins squeezed and consequently employees and all insureds are paying bigger deductibles and co-pays; consumers (i.e., voters) are experiencing the costs first hand. Government budgets are bursting and the projections of unfunded liabilities are beyond manageable. Once Healthcare Reform takes effect in 2014, there will likely be millions of currently uninsured individuals and thousands of small businesses either paying hundreds of dollars per person for medical insurance or paying penalties to the federal government. Something has to give. (Note: the penalty on businesses was recently delayed until 2015).

0%

2%

4%

6%

8%

10%

12%

14%

Rat

e

Medical Inflation vs. CPI

NCCI Medical Severity CPI – Medical Care Services CPI

1995 2000 2005 2010 2012

NCCI Medical Severity: includes the average of all medical cost per occupational accident claim for NCCI states. This index is more comprehensive as changes in utilization costs, pharmaceuticals, etc. are included. Consumer Price Index – Medical Services: one of eight subgroups within the overall CPI, this index includes the change in the cost of 1) professional services, 2) hospital and related services and 3) health insurance. Consumer Price Index: the most comprehensive Consumer Price Index includes the changes in cost of goods and services purchased by urban households.

7

Central to this transformation, we are now seeing major health groups starting to practice patient centered medicine instead of pay per procedure. They are finding ways to eliminate or streamline procedures to enhance efficiency while maintaining quality. In addition, insurers are increasingly converting to results-based reimbursement and more doctors are becoming salaried. An experiment currently being carried out by a large health plan for California public employees involves the implementation of reimbursement caps for certain scheduled procedures such as knee replacements. With these caps, the insurers will advise the patients of the cost of procedures at various medical providers and if the claimant chooses a provider that charges more than the cap, the claimant pays the overage. Early results look promising as consumers are making the decision to select less expensive procedures (Source: New York Times, June 24th, 2013). Of course, the Workers’ Compensation industry will not be able to adopt this particular cost containment mechanism, but the capping of reimbursements is creating price competition and consumer transparency that may have an indirect benefit.

The government is similarly getting into cost reduction and the Workers’ Compensation industry may be the best example of this. The majority of states have implemented fee schedules and have indexed reimbursements to Medicare rates. Based on the results of the Workers’ Compensation Research Institute’s (WCRI) Medical Price Index study (fifth edition), the benefit of these fee schedules is very apparent. The study shows the medical cost differential in 25 states by looking at the prices paid for professional services. Six of the states (the “No Fee Six”) do not have fee schedules. When ranked by

price, the top three states did not have fee schedules and all of the No Fee Six ranked within the top eight. Not only were the No Fee Six the highest priced states, but their increase in pricing from 2002 - 2012 far exceeded the study average; over this period, the increases for the No Fee Six ranged from 27% to 37% whereas the median growth for the 25 states was only 13%. As one might expect, in the fee schedule states the rate of increase in prices was strongly correlated with the changes in the fee schedules. Louisiana is a very good example of how fee schedules are effective in containing costs. From 2002 – 2012, prices and fee schedules remained relatively stable. However, the average price of pain management injections, several types of which were not covered by fee schedules, skyrocketed 82%. Visit www.wcri.org for the complete study titled “WCRI Medical Price Index for Workers’ Compensation, fifth edition.” There is another school of thought that deserves mention. Within the Workers’ Compensation industry, some believe the greater the shortage of doctors within any given population, the fewer procedures are delivered per claimant. Is it possible the emerging doctor shortage is creating a reduction in services per patient? We are not aware of any studies on this topic, but there is a certain amount of logic behind this. We expect medical inflation to continue and with an improving economy, it is possible we will we see a rebound in costs. There is a lot of low hanging fruit that can be picked to reduce costs and we expect to see a continuation and acceleration of the current cost-saving efforts as this receives more and more attention. There are solid reasons to believe medical inflation will be more subdued than previously projected. We certainly hope this will be the case.

8

GUEST ARTICLE

Are you prepared for the Workers’ Compensation “Drag Along” provision in U.S. Healthcare Reform? Point B, Inc. and Edifecs Across the U.S., healthcare providers and payors are preparing for a massive change in the way diagnoses and procedures are recorded and reimbursed. The new coding system, ICD-10, increases detail and complexity; it multiplies the number of diagnosis and procedure codes seven-fold—from thousands to tens of thousands of codes. The ultimate goal: to improve the accuracy of measuring the quality, cost, safety and efficacy of healthcare nationwide without impacting provider revenue in the near term. The interim reality: healthcare organizations face a transformational change impacting what gets measured, what people do, how they do it, who works with whom, and what people will need to do their jobs. In the near term, we anticipate that the change will be close to revenue neutral. However, planned Federal reimbursement changes in 2016 will leverage the additional detail to adjust payment levels for some diagnoses, e.g., reducing payments for others.

The “hidden mandate” for Workers’ Compensation Workers’ Compensation entities are not mandated to make the change to ICD-10. But they might as well be, since virtually everyone else in the Healthcare industry including providers and payors, will likely be adopting the new system. We call this the “drag along” effect-because providers will likely not be “dual coding” claims. In order to do business in this new healthcare environment, Workers’ Compensation insurers and their partners will need to be just as prepared for ICD-10 as everyone else.

Workers’ Compensation-related claims will be impacted by the change An evaluation of Workers’ Compensation-related codes, e.g., back, brain, burn and other codes identified by Willis Re and Point B and analyzed by health IT firm Edifecs, indicates the change in Workers’ Compensation related diagnoses and procedures is significant. For example, there are six times as many back (soft tissue) codes and more than 61 times as many spinal codes. There are more than 28 times as many burn-related codes. The growth in the number of codes is the result of these changes including additional detail such as laterality, site, severity and encounter type.

Change brings opportunity and risk The sheer volume of additional coding options presents both opportunities and risks. Best-in-class payors and their partners, e.g., bill reviewers, are preparing to benefit from the new system by revising their policies to align with it. For example, they are determining the level of coding specificity and supporting documentation is required to comply with their policies, and the specific codes are eligible for reimbursement. In addition, leading payors are planning to refine both their fraud detection rules and algorithms to benefit from new information. Ultimately, leading providers and payors are eager to identify treatments with the greatest efficacy to improve the efficiency and effectiveness of the overall system. While the groundwork for these opportunities is being laid now, the benefits will take some time to accrue. In the near term, payment processing is likely to be delayed as payors and their partners adjust their policies and claims processing rules. Best-in-class providers are anticipating and mitigating the potential cash flow impacts and even some payment volatility. While the change is intended to be net neutral in the aggregate, some diagnoses and procedures are subject to “payment shift” and may be

9

reimbursed at different levels in the new system. Each book of claims will be impacted differently, depending upon factors such as the claimant population, volume and type of claims, mix of care settings (inpatient or outpatient), and level of provider and payor or bill reviewer readiness.

Where to begin With ICD-10 on the horizon, it’s none too soon to assess how the change will impact your company and to prepare to realize benefits and mitigate risks.

Assess the right things: stakeholders, business impact and ability to execute Over the past 15 years, Point B has executed thousands of projects for more than 400 clients. Our assessment methodology focuses on the reasons most transformational efforts fail. Start by building awareness and by establishing the appropriate level of sponsorship and engagement within your organization. This is more than an “IT project,” and your executive engagement should be commensurate with the potential opportunity and risk. Understand the potential business impacts. Carefully review your business to determine how sensitive it is to the change. For example, inpatient care procedure payments, e.g., procedures typically associated with hospital stays and severe claims, are more likely to be impacted by the change than are clinic visits. Analyze your historical claims data to identify the providers and claims with the greatest risk of change. Identify the internal policies, processes and systems that are impacted by the change. And perhaps most importantly, review your entire business ecosystem: engage your providers and partners to determine their readiness levels and potential impacts to your business. Some Workers’ Compensation providers will be able to prepare for and execute the change with existing resources. Others will require support from management consultants with healthcare provider and payor expertise, such as Point B, or ICD-10 and claims processing expertise, such as Edifecs, to perform an assessment and drive change. Given the sheer volume of codes, claim scenarios and potential risk, some Workers’ Compensation providers may benefit from using ICD-10 assessment software such as Edifecs to automate much of the historical claims data analysis.

Establish a baseline, execute a plan tailored to your needs and monitor as you go Over time, the healthcare industry will have a huge amount of information on the impact of the ICD-10 conversion as providers and payors adjust the way they code and process claims. Of course, beyond industry trends, you’ll want to know how ICD-10 is impacting your organization. Identify the key metrics that will reveal the extent to which your business has been affected. For example, you may want to track claim rejections, average Accounts Payable (AP) days and volume, and turnaround times on claims, medical reviews and pre-authorizations. Establish your benchmarks before ICD-10 is implemented in October 2014, then decide how often you’ll want to recheck for impact. Rechecking monthly for the first three-to-six months may be a good cadence to begin to identify trends. Stay engaged with everyone in your ecosystem. Reach out to your high-volume providers, third-party administrators; ask how they are preparing to bill and receive billing. Everyone is in this together – and everyone benefits from communicating with each other.

Point B is an employee-owned management consulting firm that specializes in helping clients execute their strategic initiatives and deliver business results that create transformational change and competitive advantage. Learn more about Point B at www.pointb.com.

Edifecs provides healthcare software solutions that improve operational performance by streamlining the exchange of information among health plans, hospitals, and other healthcare organizations, while enabling

compliance with current mandates such as HIPAA, Operating Rules and ICD-10. Learn more about Edifecs at www.edifecs.com.

10

GUEST ARTICLE

The growth and management of physician dispensing PMSI At the beginning of the last decade, physician dispensing and drug repackaging was a little-known practice, mostly utilized by physicians who were treating patients in rural areas with limited access to retail pharmacies. Treating physicians dispense medications that are the same or similar to those dispensed by a retail or mail order pharmacy, at a premium. With time, physicians began to realize healthy profits by dispensing from their offices. The growth of physician dispensing is causing concern within the Workers’ Compensation industry as it increases costs dramatically and poses a real safety concern for injured workers.

How physician dispensing works In physician dispensing, the manufacturer sells a drug in bulk at a deep discount to a repackaging company that repackages the medication into non-standard sizes or dosages, creating a new National Drug Code (NDC) with a new Average Wholesale Price (AWP) set by the repackager. The repackaging company then sells the medication to the physician to be dispensed to the injured worker during the office visit. After dispensing, the physician usually “sells” the receivable to a physician-dispensing billing company that invoices the payor. Profit is generated at every step in this process and the cost of the drug increases. Sample physician-dispensed vs pharmacy-dispensed supply chain pricing for Carisoprodol 350mg

Legislative response Currently, Utah is the only state that completely prohibits the practice of physician dispensing. Massachusetts, Montana, New Jersey, New York and Texas severely restrict the practice, but it is still permitted if certain conditions are met. Twenty-six states have engaged in rule-making and adopted rules to regulate physician dispensing. Ten of these states (Alabama, Arizona, Connecticut, Georgia, Illinois, Michigan, Mississippi, Ohio, South Carolina and Tennessee) require usage of the underlying or original NDC for reimbursement. Without it, the invoice for the medication can be rejected. This allows reimbursement based on the medication’s original AWP instead of the inflated AWP set by the repackaging company. The other 16 states have various Workers’ Compensation rules that limit physician reimbursement or impose legal restrictions limiting a physician’s ability to dispense, such as in Florida where a physician cannot dispense a Schedule II or III medication.

11

The remaining 18 states and the District of Columbia have no legal or Workers’ Compensation limits, and therefore, the practice is not regulated.

Impact of legal or Workers’ Compensation restrictions PMSI reviewed data from 2010 – 2012 to analyze the trends in physician dispensing following legislation and rule-making. In the states that severely restrict physician-dispensing, the number of physician dispensed transactions has dropped slightly each year. Total billed amounts by physician dispensers have also been kept in check, with increases under 1%. In contrast, states that have no legal or Workers’ Compensation restrictions showed an average price per pill 111% higher than those dispensed by a pharmacy. In states with some restrictions, the average price per pill was 26% higher than those dispensed by a pharmacy. For example, average price per pill dispensed by a physician in Hawaii, where there are no legal or Workers’ Compensation restrictions, was $4.71 in 2012, compared to $1.68 dispensed by a pharmacy. In South Carolina, where restrictions are in place, the average price per pill dispensed by a physician was $2.30, and $1.90 dispensed by a pharmacy.

Solutions When physicians dispense medications in their office, they bypass the valuable “checks and balances” performed by the pharmacy and Pharmacy Benefit Manager (PBM), which allow trained pharmacists to watch for medication issues that can potentially delay an injured worker’s recovery. PMSI supports its clients’ efforts to control inappropriate physician dispensing through effective legislation and regulation to ensure safe, efficacious and cost-effective drug therapy. In jurisdictions with ineffective legislative or regulatory controls, PMSI has improved clinical outcomes by developing Specialty Networks. Contractual arrangements are negotiated with occupational medicine clinics, physician-dispensing billing companies, and individual physician practices that would historically circumvent PBM processes to ensure appropriate drug therapy. Transactions processed through PMSI’s Specialty Networks are then subject to our clinical program, MedAssess™, which provides the appropriate clinical oversight that would otherwise be lost when these entities invoice payors directly. Through Specialty Network arrangements, PMSI is able to lower costs for transactions, apply program and formulary controls, create administrative savings by reducing paper bills, provide full visibility into the injured worker’s medication history and improve the clinical efficacy of treatment, thus creating the best outcomes for injured workers.

PMSI is one of the nation’s largest full-service providers of technology-enabled, clinically driven services for Pharmacy, Critical Care, and Settlement Solutions and is devoted exclusively to the Workers’

Compensation and liability markets. For more insight on PMSI’s analytics, please download PMSI’s 2013 Annual Drug Trends Report at www.pmsionline.com/drug-trends.

12

How can we help? To find out how we can offer you an extra depth of service combined with extra flexibility, simply contact us. Begin by visiting our website at www.willisre.com or calling your local office.

The contents herein are provided for informational purposes only and do not constitute and should not be construed as professional advice. Any and all examples used herein are for illustrative purposes only, are purely hypothetical in nature, and offered merely to describe concepts or ideas. They are not offered as solutions to produce specific results and are not to be relied upon. The reader is cautioned to consult independent professional advisors of his/her choice and formulate independent conclusions and opinions regarding the subject matter discussed herein. Willis is not responsible for the accuracy or completeness of the contents herein and expressly disclaims any responsibility or liability for the reader's application of any of the contents herein to any analysis or other matter, nor do the contents herein guarantee, and should not be construed to guarantee, any particular result or outcome. This analysis has been prepared by Willis Limited and/or Willis Re Inc (“Willis Re”) on condition that it shall be treated as strictly confidential and shall not be communicated in whole, in part, or in summary to any third party without written consent from Willis Re. The views expressed in the Guest Articles are those of the authors of those articles and do not reflect any view or position on the part of Willis Re. Willis Re takes no responsibility for the content of those articles, their accuracy, completeness or any application of the contents to any particular circumstance. The reader is urged to consult its independent professional advisors in connection with the contents of those articles as well as the contents of this entire publication. Willis Re hereby disclaims all liability in connection with the Guest Articles and their contents. Willis Re has relied upon data from public and/or other sources when preparing this analysis. No attempt has been made to verify independently the accuracy of this data. Willis Re does not represent or otherwise guarantee the accuracy or completeness of such data nor assume responsibility for the result of any error or omission in the data or other materials gathered from any source in the preparation of this analysis. Willis Re, its parent companies, sister companies, subsidiaries and affiliates (hereinafter “Willis”) shall have no liability in connection with any results, including, without limitation, those arising from based upon or in connection with errors, omissions, inaccuracies, or inadequacies associated with the data or arising from, based upon or in connection with any methodologies used or applied by Willis Re in producing this analysis or any results contained herein. Willis expressly disclaims any and all liability arising from, based upon or in connection with this analysis. Willis assumes no duty in contract, tort or otherwise to any party arising from, based upon or in connection with this analysis, and no party should expect Willis to owe it any such duty. There are many uncertainties inherent in this analysis including, but not limited to, issues such as limitations in the available data, reliance on client data and outside data sources, the underlying volatility of loss and other random processes, uncertainties that characterize the application of professional judgment in estimates and assumptions, etc. Ultimate losses, liabilities and claims depend upon future contingent events, including but not limited to unanticipated changes in inflation, laws, and regulations. As a result of these uncertainties, the actual outcomes could vary significantly from Willis Re’s estimates in either direction. Willis makes no representation about and does not guarantee the outcome, results, success, or profitability of any insurance or reinsurance program or venture, whether or not the analyses or conclusions contained herein apply to such program or venture. Willis does not recommend making decisions based solely on the information contained in this analysis. Rather, this analysis should be viewed as a supplement to other information, including specific business practice, claims experience, and financial situation. Independent professional advisors should be consulted with respect to the issues and conclusions presented herein and their possible application. Willis makes no representation or warranty as to the accuracy or completeness of this document and its contents. This analysis is not intended to be a complete actuarial communication, and as such is not intended to be relied upon. A complete communication can be provided upon request. Willis Re actuaries are available to answer questions about this analysis. Willis does not provide legal, accounting, or tax advice. This analysis does not constitute, is not intended to provide, and should not be construed as such advice. Qualified advisers should be consulted in these areas. Willis makes no representation, does not guarantee and assumes no liability for the accuracy or completeness of, or any results obtained by application of, this analysis and conclusions provided herein. Where data is supplied by way of CD or other electronic format, Willis accepts no liability for any loss or damage caused to the Recipient directly or indirectly through use of any such CD or other electronic format, even where caused by negligence. Without limitation, Willis shall not be liable for: loss or corruption of data, damage to any computer or communications system, indirect or consequential losses. The Recipient should take proper precautions to prevent loss or damage – including the use of a virus checker. This limitation of liability does not apply to losses or damage caused by death, personal injury, dishonesty or any other liability which cannot be excluded by law. Acceptance of this document shall be deemed agreement to the above. Risk Management Solutions (RMS) “The technology and Licensed Data used in providing this information is owned by Risk Management Solutions, Inc. and its licensors, and is based on the scientific data, mathematical and empirical models, and encoded experience of earthquake engineers, wind engineers, structural engineers, geologists, seismologists, meteorologists and geotechnical specialists. As with any model of complex physical systems, particularly those with low frequencies of occurrence and potentially high severity outcomes, the actual losses from catastrophic events may differ from the results of simulation analyses. Furthermore, the accuracy of predictions depends largely on the accuracy and quality of the data input by the user. This information is confidential and may not be shared with any third party without the prior written consent of both Willis Group Services Limited and Risk Management Solutions, Inc. Furthermore, this information may only be used for the specific business application specified by Willis Group Services Limited and for no other purpose and may not be used under any circumstances to support development of or calibration of a new or existing product or service offering that competes with Risk Management Solutions, Inc. THIS INFORMATION IS PROVIDED “AS IS”, AND RISK MANAGEMENT SOLUTIONS, INC. DISCLAIMS ALL WARRANTIES, WHETHER EXPRESS OR IMPLIED, WITH RESPECT TO THE INFORMATION, INCLUDING BUT NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT SHALL RISK MANAGEMENT SOLUTIONS, INC. OR WILLIS GROUP SERVICES LIMITED BE LIABLE FOR INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES OF ANY KIND ARISING FROM ANY USE OF THIS INFORMATION.

Contact information Simon David Sam Dutcher Brian Ingle Executive Vice President Executive Vice President Executive Vice President Willis Re Inc Willis Re Inc Willis Re Inc 525 Market Street, Suite 3400 525 Market Street, Suite 3400 One World Financial Center San Francisco, CA 94105 San Francisco, CA 94105 200 Liberty Street +1 415 955 0277 +1 415 955 0279 New York, NY 10281 simon.david@willis.com sam.dutcher@willis.com +1 212 915 8446 brian.ingle@willis.com