Costing and the pursuit of value in healthcare

Costing and the pursuit of value in healthcare Robert Kaplan value masterclass shaping healthcare finance

Contents Introduction 3 Outcomes 3 Costing 5 Value-based payment 8 Questions and answers 10 Further reading 10 Healthcare Financial Management Association June 2015 All rights reserved. The copyright of this material and any related press material featuring on the website is owned by the HFMA. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopy, recording or otherwise without the permission of the publishers. Foreword Delivering value has to become the focus for clinicians and managers alike as they look to deliver high-quality healthcare services to meet the needs of a growing and ageing population. We can t simply focus on services with no regard for the costs. Similarly, simply looking to bear down on costs with no regard for the impact on care and outcomes is likely to be counterproductive and is simply not sustainable. So the theory of bringing both sides quality and costs into the decision-making process for all decisions about health and healthcare services makes sense. But the trick is turning this theory into practice. Recognising that value is important is only part of the battle. What the NHS needs and other healthcare systems are practical tools that enable them to take a rounded view of the value when assessing current services, service developments and weighing up alternative treatment approaches. In the US, two professors Michael Porter and Robert Kaplan have championed the adoption of value measured as outcomes over costs as the real goal for healthcare systems. Their work has considered how you identify and measure outcomes that are important to patients. Professor Kaplan in particular proposes time-driven activity-based costing as a mechanism for understanding the important cost denominator. It was therefore fitting that the HFMA Healthcare Costing for Value Institute marked its launch event with a masterclass with Professor Kaplan. It was fascinating to hear the theory outlined by such a passionate advocate and a lively question and answer session proved that this is a topic on the minds of clinicians and finance managers across the NHS. This briefing summarises Professor Kaplan s session at the Institute s masterclass. I hope you find it useful. The HFMA Healthcare Costing for Value Institute will continue to support practitioners as they look to make a reality of value-based decision-making in the NHS. Paul Briddock HFMA policy and technical director

3 Introduction The only way to meet the demands facing modern healthcare systems is to pursue the delivery of value and that means a detailed understanding of outcomes and costs at medical condition-level and across whole cycles of care. This was the clear message from Professor Robert Kaplan of the Harvard Business School when he gave the keynote speech to an HFMA Healthcare Costing for Value Institute masterclass in April. Professor Kaplan has been a leading proponent of the need to focus on value alongside fellow Harvard professor Michael Porter. Speaking via a videoconference link, Professor Kaplan told the audience: In the US, we have a forgiving and generous reimbursement system based on fee for service. The goal is to drive revenues to cover expenses. But that era is coming to an end or has come to an end not just in the US. And we are looking for new ways to deliver medical solutions to patients while containing the costs that have inexorably risen over the last 20 years. He described the value approach as providing an optimistic framework, but said that health systems had little choice. This is the only way we can go in transforming healthcare systems in a way where we don t ask healthcare workers to take a pay cut, limit access through rationing or devote more and more GDP to healthcare when there are other pressing needs for society. The value approach offered an opportunity for better outcomes and higher capacity, translating into lower unit costs, he said. Professor Porter had defined value as the ratio of DIAGRAM 1: VALUE DEFINITION Value = Health outcomes Costs of delivering the outcomes Outcomes are the full set of patient health outcomes over the cycle of care Costs are the total costs of resources used to care for a patient s condition over the care cycle healthcare outcomes and the costs of delivering those outcomes. Professor Kaplan said he enthusiastically endorsed this view. His role had been to provide a more detailed view on the costing side, in particular proposing time-driven activity-based costing as the most appropriate way to determine this value denominator. Professor Kaplan said that even with agreement about the definition, the unit of analysis needed to be agreed. The problem is that most systems are looking at the wrong one the hospital or healthcare provider, he said. When you look at this level, what you are left with are somewhat generic outcomes. That is where we are in the US. They say we want to improve outcomes, but they measure outcomes at the institution level so they are limited in choice. He listed a number of typical measures such as patient safety, avoiding infections, avoiding readmissions and length of stay. Clearly they are important but they are only quasi outcomes. They are not granular enough to get what we want, he said. Instead, value had to be measured at the medical condition level for example, lung cancer, diabetes, hip problems or congestive heart failure and you need to look at the outcome of the whole cycle of care and not just part of it. The first step is to identify and understand this whole cycle of care for different conditions which can be achieved by process mapping. Ideally, the cycle should be looked at from when a patient first enters the health system at the primary care stage. However, practicalities might mean looking at cycles that start once a decision is made to do surgery or undertake some other intervention. And to make the value framework operational for long-term conditions, where the cycle of care is ongoing, it might be necessary to choose an appropriate period of care. Professor Kaplan identified three building blocks for a value-based healthcare system: l Measure and communicate outcomes by medical condition l Measure and improve costs by medical condition l Develop bundled payments to compensate providers for treating medical conditions. Getting the first two of these in place, measuring outcomes and costs for patients at medical condition level, would provide the opportunity for the third. If you don t change the payment system to align with value, you won t get all the benefits that you can get, he said. Outcomes Professor Kaplan said that to understand outcomes, you also need an understanding of the patients at the outset of the treatment. This would include basic features such as age and weight, but also cover the presence of comorbidities whether the patient has diabetes, for example.

4 Costing and the pursuit of value in healthcare Avoiding defects is not the same as putting a cancer into remission Or are there any other relevant factors does the patient smoke or have substance abuse or alcohol problems, perhaps? The I in your PLICS [patient level information and costing systems] will be increasingly important to understand these initial conditions, said Professor Kaplan. He added that patient experience was also an important part of the overall patient outcome. He suggested there was a tendency to measure what you can rather than what you need to measure. This had led to a focus on the process measures or quasioutcomes he had mentioned earlier infection rates and readmissions, but also staffing ratios and adherence to guidelines. He stressed again that these were all important measures, but that they were insufficient to measure value. In the US, there are currently plans to make half of all Medicare payments (the system that covers the over-65 year-olds) based on the quality of care provided rather than quantity by 2018. However, to date, Professor Kaplan said that 90% of the outcome measures developed were really process measures and even these measures can take up to three years to be approved as usable measures. Professor Kaplan said it was necessary to recognise the difference between process measures and outcome measures. Avoiding defects is not the same as putting a cancer into remission, he said. He highlighted work by his colleague Professor Porter that identified three categories of outcomes: l Tier one would reflect the health status achieved or retained survival (mortality rates) or the degree of health/recovery (achieved clinical or functional status). l Tier two would measure the patient s experience during the care cycle (how long did the treatment take?) and the disutility of the process (were there any care-related pain, complications or the need for reinterventions?). l Tier three would look at the sustainability of the patient s health (what was the long-term TABLE 1: PROSTATE CANCER OUTCOMES Tier 1 Survival Five-year survival rate Degree of recovery/health Continence (one year) Erectile function (one year) Other quality of life Tier 2 Time to recovery or return to normal activities Time to diagnosis Time to treatment Length of inpatient stay Time to return to work Disutility of the care or treatment process Bleeding Thrombosis Short term continence (one week, three months) Short term erectile function (three months) Tier 3 Sustainability of recovery or health over time Biomechanical recurrence Metastatic progression Long-term consequences of therapy Radiation-induced complications of intestine, bladder, bones, skin

5 clinical/functional status?) and the consequences of the treatment. Professor Kaplan provided examples of outcome measures for prostate cancer that would fit into this three-tier model (see table 1). Once outcomes are measured in this more meaningful way in particular taking account of outcomes that are really important to patients it provides opportunities to understand the value delivered by different approaches. Professor Kaplan highlighted work at a specialist prostate cancer centre, the Martini Klinik in Hamburg. The clinic started to measure outcomes in 1994, initially on an Excel spreadsheet, although this has since moved onto an electronic patient record. Quality of life and outcomespecific surveys were filled out pre-operation, on discharge and post discharge at three months, one year, two years and three years. Because patients are reluctant to return to the clinic to assess outcomes after a year, the clinic has moved to web-based surveys. By 2013, the clinic was conducting 1,200 surveys a month with a 90% return rate (after multiple phone reminders) and is now at the stage of having a database of some 20,000 prostate cancer patients. The outcomes are used for learning and improvement, with data on each surgeon s outcomes shared with all urologists every six months. There is also an annual public report reporting survival rates and outcomes on continence and erectile function. Five-year survival rates for the clinic are in line with the average for German hospitals (95% compared with the national average of 94%). But that is where the similarities end. The rates for severe erectile dysfunction and incontinence key outcomes for patients are nearly five times better for patients at the Martini Klinik. Another example of outcome measurement good practice can be found at the Hoag Orthopaedic Institute in California. The facility specialises in orthopaedic surgery and routinely measures detailed outcomes, including functional outcomes. It again uses a webbased collection system. An annual outcomes report is published and, according to Professor Kaplan, the organisation basically undertakes no further marketing the results talk for themselves. Although identifying outcomes by medical condition might seem daunting, Professor Kaplan said much of the work was now undertaken centrally. For example, non-profit body the International Consortium for Health Outcomes Measurement (www.ichom.org) began a programme to support clinicians to measure outcomes back in 2013. Outcomes for four medical conditions were published in 2013 (coronary heart disease, lower back pain, cataracts and localised prostate cancer). These were followed in 2014 by eight further conditions, meaning an estimated 35% of the burden of disease has now been covered. The point is you don t have to reinvent the wheel, said Professor Kaplan. Costing Having covered the numerator of the value equation in some detail, Professor Kaplan moved onto the equation s denominator costs. Again he stressed the importance of looking at costs across the whole care cycle (or the same cycle used to examine outcomes). Getting clinicians interested in costs can seem challenging. But their engagement is vital to ensure that costs are accurately calculated and then to use the resulting costs, alongside outcome data, to drive improvement. Professor Kaplan said that clinicians were typically mission driven, seeing margin and costs as administrators territory. Professor Kaplan said clinicians were mission driven, seeing margin and costs as administrators territory. But, he said, mission and margin were compatible

6 Costing and the pursuit of value in healthcare Who does what determines how efficient we are However, he said that the two sides mission and margin were completely compatible. We have to find a way to be mission-oriented in a way that preserves the financial margin, he said. After five years of doing this I can assure you that clinicians in fact enjoy discussions about cost but only if the way we measure makes sense to them, he said. So to understand costs for the care provided for a particular condition, it is back to the process maps. But this time we ask two additional questions: who does [each step] and how long does it take? Professor Porter said. Talking through the basic approach to time-driven activity based costing, Professor Kaplan showed examples of process maps that also identified the personnel and equipment involved in each step and the time taken (see diagram 2). The next step involves identifying the total costs associated with having these different personnel available to treat patients and their total capacity how much time these personnel actually have available for treating and caring for patients. This enables cost/minute rates to be calculated and assigned to each of the process steps for the relevant DIAGRAM 2: PROCESS MAP Colour coded to show the staff involved and the time taken for each step check-in Waiting room Exam room X-ray room Education room check-out arrives checks in fills out paperwork 2 5 assessment Take X-rays 7 Bring patient to education room checks out 2 2 Yes 85% Review and record patient info 5 waits for appt 20 Need to X-ray? Process and annotate image 5 watches video 20 departs No 15% Staff key Office assistant Physician assistant X-ray technician Surgeon Scribe RN Preview patient info before appt waits for surgeon 20 Yes 60% Discuss with patient and develop Dictate notes and consult with staff as 7 care plan 10 needed 8 Discuss surgery and answer questions 15 No 40% Transcribe notes having a TKA? 8

7 personnel. A total cost for each process is calculated and then the costs of all the process steps in a care pathway can be added up. As well as providing accurate estimates of the costs involved in treating different conditions, the approach provides a good way to identify potential cost improvements if personnel were used in different ways. Professor Kaplan said there are often 10:1 variations in the personnel capacity rates for the different people involved in the care cycles. Who does what determines how efficient we are, he said. In the US we have $6 a minute surgeons doing work that 60 cent a minute assistants could do equally well. This is an inefficient use of resources, but you don t see this unless you get to this calculation. Time-driven activity-based costing was the way to break down the wall, said Professor Kaplan, enabling finance professionals and clinicians to hold productive discussions about where procedures could be standardised or whole steps missed out to deliver consistent, high-quality outcomes at lower costs. said. Space is not that expensive even operating room space. He suggested that the cost per minute of having an operating room available was typically 40-50 cents. But a well-trained surgical team costs about $20/minute. What do you want maximum utilisation of your $20/ minute resource or your 50 cents/ minute resource? It s a no brainer. Perhaps the missing tool for clinicians and finance practitioners is bringing all the outcome and cost information together in a format that can be easily analysed. Professor Kaplan says the solution is to use radar or spider web charts (diagram 3). The further from the origin an outcome or cost is plotted, the better the outcome or cost with cost being plotted as 1/cost. Spider web charts bring all the outcome and cost information together in a format that can easily be analysed DIAGRAM 3: ANALYSIS OF OUTCOMES AND COSTS ON SINGLE PAGE Brachytherapy treatment for prostate cancer Costs Acute complications Professor Kaplan described a joint replacement study across 30 hospital sites that revealed a huge 1.7 times cost variation from the 10th percentile to the 90th percentile. This variation was after standardising the cost of the individual staff types and so represented real variations in process. The opportunity for improvement is enormous, he said. Some surgeons were doing 10 joint replacements a day and other surgeons did three a day. The time in the operating room was the same, but the high-productivity surgeons had access to two operating rooms this was the big driver [of the variation]. He said there was a myth that operating rooms are expensive. But this turns out not to be true, he Survival and disease control reported health status

8 Costing and the pursuit of value in healthcare The power comes when you compare, say, four different ways to treat prostate cancer, he said (see diagram 4). In this example, while outcomes across all the outcome measures are similar, there is a very visible significant cost benefit from the brachytherapy approach. This is not the whole story because other issues also need to be considered. For example, organisations or health systems may want to be able to offer a choice of treatments to patients and this might override any considerations of cost. However, the value approach at least means that all the information is on the table. The same all data on a page approach could be used in a number of different contexts. You could put this across different surgeons or different facilities for the same treatment, explained Professor Kaplan. It is a very powerful way to display value covering both outcomes and costs. Value-based payment The final step, according to Professor Kaplan, is to devise a payment system that drives organisations to focus on value. DIAGRAM 4: ANALYSIS OF DIFFERENT APPROACHES TO TREATING PROSTATE CANCER Brachytherapy Photon (IMRT) Proton (IMPT) Radical Prostatectomy Sexual function (EPIC) 1/Cost Urinary incontinence (EPIC Cure brfs (%) Urinary bother (EPIC) Bowel function (EPIC)

9 Current payment mechanisms around the world typically use one of three different approaches: l Fee for service l Global provider budgets l Global capitation budgets. He suggested that all these systems were flawed, either driving volume rather than value or transferring risk in an unsustainable way. Instead, he said, a bundled payment approach was needed, with a single payment given for treating a condition over the full cycle of care. And he added that some of the payment needed to be at risk based on outcomes, with the payment and outcome targets risk-stratified by the complexity of a provider s patient population. Professor Kaplan outlined four fundamental components in a value-based bundled payment: l Covers care for a medical condition, not for a procedure or treatment episode l Contingent on condition-specific outcomes that matter to patients l Negotiate the payment based on the cost of efficient and effective care, not past charges l Specifies limits of responsibility for unrelated care needs and catastrophic events. Again, Professor Kaplan offered an example where a Swedish health system had used bundled payments to drive improvement in value. Stockholm County had had a problem with joint replacement surgery, with patients frequently on waiting lists for up to two years. Earlier approaches, such as offering providers more money to do more procedures had had little impact and out-of-county providers had to be paid to help tackle the backlog. The health authority adopted a bundled payment approach, giving a fixed fee to cover physician fees, all other personnel costs, occupancy in hospital, drugs, implants, tests and supplies. Outpatient rehabilitation and additional inpatient rehabilitation were kept separate. The cycle of care covered included the pre-op consultation, surgery, inpatient recovery and one follow-up visit for the two lower risk surgery levels of a total of four risk levels. With a twofold variation in the prices charged by private hospitals under the previous reimbursement system, a relatively low price was set based on a price voluntarily entered into by one private provider, which had delivered good outcomes. In the first year, there was an increase in activity one of the key goals but this was accompanied by a 17% drop in costs and a 33% fall in the complication rate. By 2011, the surgery queue had disappeared. Other health systems are now looking to follow this example. In the US, the University of Texas MD Anderson Cancer Center and United Healthcare have launched a pilot to explore a new cancer care payment model for head and neck cancers. The system will focus on quality patient care and outcomes and is believed to be among the first to use bundled payments in a large and comprehensive cancer unit. The pilot follows work with the Harvard Business School. That s our vision. We d love to be there in five years time for all major conditions, said Professor Kaplan. But we have to start now. l Current payment mechanisms are flawed, either driving volume rather than value, or transferring risk in an unsustainable way

10 Costing and the pursuit of value in healthcare FURTHER READING What is value in healthcare? Having earlier set out the concept of value and health outcome measurement, Professor Michael Porter further develops his ideas in this New England Journal of Medicine article from December 2010. tinyurl.com/95zjzkc The big idea: how to solve the cost crisis in health care Perhaps the most frequently referenced article on healthcare value, this September 2011 Harvard Business Review article brings Professors Kaplan and Porter together to restate the value proposition and set out the argument for time-driven activity based costing tinyurl.com/nhbmqqh The strategy that will fix healthcare The value case is remade in this October 2013 Harvard Business Review article by Professor Porter and Thomas Lee tinyurl.com/pp6gs9y How not to cut healthcare costs Professor Kaplan teams up with Derek Haas in this November 2014 article to outline five common mistakes made when hospitals try to cut costs tinyurl.com/q6d49h3 Questions and answers Q With tens of thousands of different procedure and diagnosis codes and thousands of healthcare resource groups, how many outcome measures need to be developed? Professor Kaplan said that it was important to get the unit of analysis right and the detailed level defined by procedure codes or healthcare resource groups was not the right level. If we did 70 medical conditions, we would pick up 80% - 90% of the medical conditions you face and 95% of the spend. And that is where we ve got to go. Again he highlighted the work by the International Consortium for Health Outcomes Measurement. Its website claims that under current plans, it will have published 50 standard sets by 2017, covering more than 50% of the global disease burden. Q At what point should quality-related payments be made? Professor Kaplan accepted that outcomes sometimes needed to be measured over several years. However it was not practical to hold back payment until final outcomes could be measured. You can t have 10-year holdbacks, he said. So I d say: make 90% of the payment at the time and a 10% holdback paid at the end of the year based on the outcomes. Q What link should there be between costs and price? Professor Kaplan accepted this was not straightforward. He pointed to the Stockholm County joint replacement programme, where the price had been set independent of costs. Instead, the health authority had chosen a relatively low price already used in a contract with a provider that had been entered into voluntarily (demonstrating it was achievable) and where good outcomes were being delivered. However, he suggested that a fixed price for example, for three years could provide some certainty for providers. You don t want to look at costs and ratchet the price down each year because then there is no return for the innovation that goes on and they don t have the margin they need to expand their business, he said. I want people with great outcomes and good margins to do more. So let s choose prices based on where we are today and find out who can survive and prosper.

About the HFMA The Healthcare Financial Management Association (HFMA) is the professional body for finance staff in healthcare. For more than 60 years, it has provided independent and objective advice to its members and the wider healthcare community. It is a charitable organisation that promotes best practice and innovation in financial management and governance across the UK health economy through its local and national networks. The association also analyses and responds to national policy and aims to exert influence in shaping the wider healthcare agenda. It has a particular interest in promoting the highest professional standards in financial management and governance and is keen to work with other organisations to promote approaches that really are fit for purpose and effective. Healthcare Costing for Value Institute The aim of the Healthcare Costing for Value Institute is to drive improvements in the quality of cost and business information used within the NHS to allow for more robust and value-based decision-making. The NHS is now at a point where it is recognising the importance of good costing, the need for highly trained costing staff and the possibilities of costing for value. The Institute is in an ideal position to help organisations meet these needs and move the agenda forward. To find out more about becoming a member of the HFMA Healthcare Costing for Value Institute, please contact James Blackwell, business development manager T: 0117 938 8446 E: james.blackwell@hfma.org.uk HFMA 1 Temple Way Bristol BS2 0BU T 0117 929 4789 F 0117 929 4844 E info@hfma.org.uk www.hfma.org.uk Healthcare Financial Management Association (HFMA) is a registered charity in England and Wales, no 1114463 and Scotland, no SCO41994. HFMA is also a limited company registered in England and Wales, no 5787972. Registered office: 110 Rochester Row, Victoria, London SW1P 1JP.