Medicare Cost Data Analysis

1 Executive Summary

Data analysis was performed on the recently published Medicare Provider Utilization and Payment Data for (a) Physicians and Other suppliers and (b) Inpatient. A lot of news articles have been written in the past about the huge variability in billed charges among regional hospitals and how some physicians are getting reimbursed millions of dollars by Medicare. The current data analysis tries to shine a broader light on the data and provide the general public a narrative of what the published data tells us. The data analyses will be expanded over time as time permits and based on comments from readers.

The data analysis found that at least 60 doctors have had Medicare reimbursements above $5 million in calendar year 2012. The top doctor had $21 mil in Medicare reimbursements. All of these reimbursements may have been legitimate but then it begs the question of how a single doctor can be so productive and efficient. Are these the super-doctors that the medical field should emulate? Are the taxpayers and retirees getting full value for their money at these levels of reimbursements?
Ophthalmology is the second highest doctor category in terms of Medicare reimbursements, after Internal Medicine. It is even higher than Cardiology. What are the common ophthalmological procedures done on retirees other than cataract operations? If ophthalmological procedures are that expensive, then we are not hearing enough ads in the general society about preventive measures on eye health like the way we hear about healthy heart.
The number of physicians billing Medicare in each state is more or less proportional to the general population of the state. California, the most populous state, tops the number of Medicare doctors with 71,130 doctors. Even though Florida has more people older than 65 compared to New York or Texas, Florida still only comes in fourth (behind New York and Texas) in terms of the number of doctors servicing Medicare population. The Florida doctors could be seeing more Medicare population per doctor compared to New York or Texas.
Medicare paid $9.7 billion to Clinical Laboratories and Ambulatory services across the nation in CY2012. The top two laboratory service providers are Lab Corp and Quest Diagnostics.
The total hospital inpatient services payments made by Medicare in the provided data only added upto $57 billion. It is less than half the amount reported by the Kaiser family foundation on its website. The data only covers the top 100 most frequently billed discharges.
Major Joint replacement topped the Hospital category in terms of Medicare reimbursements, at $5.2 billion. Septicemia or severe Sepsis came second at $4.4 billion. The reimbursements on heart failure is only $3 billion. There are lot of societal ads on keeping heart healthy but not enough on keeping limb bones and knee caps healthy. Interestingly Medicare pays on average $12,500 for a joint replacement compared to $9,500 for heart failure. Joint replacements are more expensive and more retirees are getting it done compared to heart failure.
Some of the arguments heard in news articles on controlling Medicare costs revolve around how a lot of expenses in Medicare is around the final stages of older people supporting them on ventilators. The payments by Medicare on respiratory support with ventilator systems amount to around $1.9 bil in CY2012. It is a cost but a smaller cost compared to the spending on other ailments.
New York Presbyterian hospital topped the list of hospitals in terms of Medicare payments at $255 million, followed by Florida hospital in Orlando and MonteFiore Medical Center in Bronx.

Images of the data analysis are provided in the analysis below. The data analysis was prepared using Tableau public. The source of the data analysis prepared by me is available here:
https://public.tableausoftware.com/views/MedicareDoctorsDataAnalysis/MedicarePhysicianpaymentanalysis?:embed=y&:display_count=no

https://public.tableausoftware.com/views/HospitalMedicarepayments/MedicarepaymentstoHospitals?:embed=y&:display_count=no

Data Drills can be done at these links.

2 Medicare Cost Analysis

This year around April 2014, Medicare published Medicare Provider Utilization and Payment Data for (a) Physicians and Other suppliers (b) Inpatient (c) Outpatient. The Physicians and Other suppliers data cover the payments to Doctors and Other suppliers (Laboratories, Ambulatory services etc). The inpatient data include hospital-specific charges for the more than 3,000 U.S. hospitals that receive Medicare Inpatient payments for the top 100 most frequently billed discharges. The outpatient data include estimated hospital-specific charges for 30 Ambulatory Payment Classification (APC) Groups paid by Medicare. They have been published for calendar year 2012 and 2011. The data set contains a wealth of information and I decided to dig into it using Tableau Public. I want to see if I can find some useful information out of this data that can provide some valuable input to the ever running debate on health care cost inflation. I did not analyse the Outpatient data. I am currently rambling on data analysis on this data set. I am moving forward without an end in mind right now – I hope I will get more insights to raise useful questions on seeing the results of the various data analyses. All the analyses have been done using the calendar year 2012 data. I will at some point start comparing 2012 against 2011 to look at trend on YoY.

2.1 Medicare General Information

Medicare is the federal health insurance program for people who are 65 or older, certain younger people with disabilities, and people with End-Stage Renal Disease (permanent kidney failure requiring dialysis or a transplant, sometimes called ESRD). Medicare has 4 parts and the different parts of Medicare help cover specific services:

Medicare Part A (Hospital Insurance)

Part A covers inpatient hospital stays, care in a skilled nursing facility, hospice care, and some home health care.

Medicare Part B (Medical Insurance)

Part B covers certain doctors' services, outpatient care, medical supplies, and preventive services.

Medicare Part C (Medicare Advantage Plans)

A Medicare Advantage Plan is a type of Medicare health plan offered by a private company that contracts with Medicare to provide you with all your Part A and Part B benefits.

Medicare Part D (prescription drug coverage)

The data released covers Medicare Part A and Part B payments. There is a good article on the Kaiser Foundation listing the various news articles that covered the Medicare Physician billing data. http://www.kaiserhealthnews.org/daily-reports/2014/april/09/medicare-billing-data.aspx. One nugget of info is like this: Reimbursements to doctors who provide Medicare services in 2012 ranged from nearly $21 million to a single Florida ophthalmologist to the $27,000 for the average anesthesiologist, according to the first look at government payment data in 35 years.

One of the articles quote that one of those ophthalmologists, Salomon Melgen, who took in more than $20 million from Medicare in 2012 was raided by the FBI and Department of Health and Human Services. I am not sure if he was raided before the data release or after the data release. I am going to go over data analysis in this blog that has not been covered by the news articles quoted in the Kaiser’s web page.

http://kff.org/medicare/fact-sheet/medicare-spending-and-financing-fact-sheet/. Medicare benefit payments totaled $583 billion in 2013; roughly one-fourth was for hospital inpatient services ($140 bil), 12% for physician services ($70 bil), and 11% for the Part D drug benefit ($64 bil). Another one-fourth of benefit spending was for Medicare Advantage private health plans covering all Part A and B benefits. In 2013, spending on Medicare accounted for 14% of the federal budget. Medicare also plays a major role in the health care system, accounting for 20% of total national health spending in 2012, 27% of spending on hospital care, and 23% of spending on physician services.

1.1 Physicians and Other Suppliers Data Analysis

There are actually two types of data sets that have been provided. The first set provides very detailed physician payments (by each payment record) and has more than 9 million records (1.7 GB in size). The second set provides a summary of physician payments over the calendar year, the number of unique patients they have seen, number of procedures ordered, type of provider, total medicare payments etc. I have used the summarized data as it is more manageable to do data analysis using Tableau Public.

The total number of providers in this data set is 880,644, of which 523,086 are male doctors, 302,026 are female doctors and the other 55,532 providers are laboratories, ambulatory service providers. The total Medicare payment made to the providers in this data set is $77 bil in CY2012. This data set has redacted any doctor who had 11 or less unique beneficiaries.

1.1 Hospital Payments Data Analysis

The Hospital payments data gave the hospital provider name details, the average Medicare payments, average Covered charges, average total payments. The average Covered charges indicate how much the hospital billed for that diagnostic procedure on average and it is usually 3.5 to 4 times higher than what Medicare paid. There are big variations in the average covered charges from hospital to hospital and this has been covered in numerous news articles in the past.