What do the numbers really mean?
The healthcare debate often turns on numbers, but statistics are easy to distort or misunderstand. We’ve talked with experts to dig into a few numbers. We will be adding interviews on more statistics in coming weeks. If there are other numbers you’d like to know more about, let us know.
The CDC estimates that there are 1.7 million nosocomial or hospital-acquired infections and 99,000 associated deaths in the United States each year. Of those:
- 32%: Urinary tract infections
- 22%: Surgical site infections
- 15%: Pneumonia (lung infections)
- 14%: Bloodstream infections
Q3 talked with David Perlin, director of the Public Health Research Institute of the New Jersey Medical School, to better understand the numbers. His research includes work on two important areas of healthcare-acquired infections: bloodstream infections and fungal infections.
Q: What should we make of these numbers?
The figure 1.7 million is probably a pretty good estimate. It may be low. At this stage we have to keep asking "Why do we still sustain 100,000 deaths a year due to hospital-acquired infections?" We should be able to do better.
The public should have the information, especially on staph infections or MRSA. People should be allowed to make an informed decision about where to get treated and what the possible risks are. But a lot of this has to do with the nature of the hospital environment you’re in.
If you’re at a hospital that deals with high-risk patients such as transplant patients or where patients are resident for long periods of time, you have to expect you’re going to see a higher rate of opportunistic infections as opposed to a lower -risk population that’s in and out — it’s something that you really have to balance and put in proper perspective. So just publishing the numbers, I think, is a bit misleading.
The assumption that people get infections in the hospital because of lax procedures, of poor infection control is to some degree true. That is the way infections like staph spread, but it doesn’t fully account for the fact that individuals are already colonized with many organisms. Fungal infections are already in your body, and you develop systemic infection when you become severely immunosuppressed. It’s not from an exogenous source. It’s from your own body.
But there’s no question there are ways to improve. All you have to do is look at the Netherlands. You’re not allowed into the main part of the hospital until they put you through a triage center. They look to see if you have, for example, S. aureus. If you do, you then get treated for that. They have remarkably low rates of staph infections. There are now pilot programs in the U.S. where they did something very similar and were able to significantly reduce the number of staph infections.
Overall, the numbers have not changed in quite a while. It may come down to the fact that hospital practice has not changed very much in the last 15 or 20 years. The way people are managed within hospitals, within intensive care units, or as they are moved from one ward to another has not changed. We really need to reassess all aspects of healthcare delivery as it may impact infections. And I don’t think we’ve really done that yet.
Interview conducted and edited by Ted O’Callahan
Source for statistics: CDC
Costs of Hospital-Acquired Infections in Pennsylvania in 2006*
$175,964: Mean hospital charge for patients with a hospital-acquired infection
$33,260: Mean hospital charge for patients without infections
$79,670: Median hospital charge for patients with a hospital-acquired infection
$18,538: Median hospital charge for patients without infections
12.3%: Mortality rate for patients with a hospital-acquired infection
2.1%: Mortality rate for patients without a hospital-acquired infection
*Hospitals do not typically receive full reimbursement of charges; on average statewide in 2006, hospitals were paid 27% of established charges.
Joe Martin, the director of communications and education for the Pennsylvania Health Care Cost Containment Council, explains the significance of the study.
Q: Could you explain the key numbers in the research?
Pennsylvania hospitals had 1,574,170 total cases in calendar year 2006. Of those, 30,237 were identified by hospitals as patients who contracted an infection after they were admitted to the hospital. The average mean charge for those patients was $175,964. The 1,543,933 cases that didn’t have a hospital-acquired infection carried an average charge of $33,260.
We don’t know the degree to which infections are driving length-of-stay or mortality issues. What we do know is that Patient A had an infection, Patient B didn’t, and the differences in mortality, length of stay, and the financial impact are fairly substantial. We're not trying to mislead people into thinking that it is all attributable to those infections. The underlying drivers and other factors at play are fertile areas for research. Are certain categories of patients more susceptible than others? Do certain surgical categories carry more risk? What we’re trying to do in our report is to make some simple observations about this issue.
Because some cases are just off the charts — a patient that’s in the hospital for 120 days at a charge of $1.8 million — and extraordinary outliers can have a big impact on the average, our board felt that we wanted to report both mean and median figures.
I’m going to speak very generally, but the majority of these infections are considered preventable through proper hand hygiene, elevating pneumonia patients, sterile precautions, the appropriate timing in the administration of pre-surgery antibiotics, and a whole host of process measures as opposed to risk factors such as how sick the patient is, how old the patient is, or if they’re immunocompromised.
Q: What should people be taking away, when they see these numbers?
It depends on who you are. If you are a patient, what you need to take away is that you or a relative or an advocate need to be as vigilant as possible about prevention during your hospitalization. If you are somebody who is purchasing healthcare benefits for a company or labor union health-and-welfare fund, you ought to be thinking about preventable expenses on the bottom line, and engaging in dialog with insurers and with hospitals and physicians about how to drive these infections down, not only because you want to drive your costs down, but because it’s better for your members and employees. If you’re a provider, this gives comparative benchmarks and other valuable information which can help them drill down to identify process problems and improve the quality of care. And if you’re a policymaker, as we’ve seen in Pennsylvania, it has helped to drive broader quality improvement efforts, where the governor and legislative leaders passed a piece of legislation that involves a series of quality improvement activities through not just our agency but the Health Department and the Patient Safety Authority, in order to encourage and assist providers as much as possible in trying to root these infections out and prevent them. Long answer, but there isn’t one single audience for this.
Interview conducted and edited by Ted O’Callahan
Administrative Costs in Healthcare as of 1999
31%: Administrative costs as a portion of U.S. healthcare expenditures
16.7%: Administrative costs as a portion of Candaian healthcare expenditures
11.7%: Average administrative overhead for private insurers in the U.S.
6.8%: Administrative overhead for Medicaid
3.6%: Administrative overhead for Medicare
1.3%: Administrative overhead for Canada's national health insurance program
26.9%: Percent of U.S. physicians’ gross income going to administrative costs
16.1%: Percent of Canadian physicians’ gross income
24.3%: Average administrative costs in U.S. hospitals
12.9%: Average administrative costs in Canadian hospitals
Steffie Woolhandler, associate professor of medicine at Harvard Medical School and co-founder of Physicians for a National Health Program, talked with Q about her work on administrative costs, originally published in the New England Journal of Medicine.
Q: Your study reported that 31% of the healthcare dollar in the U.S. goes to administrative costs. What surprises did you have in gathering this data?
From my own clinical practice I’m aware of all the people pushing papers around and sitting in front of computer screens. But when you add it all up and come up with a figure like 31% of health spending, the magnitude of the total spending on paperwork in the United States is pretty shocking.
On a per capita basis, the amount in the U.S. is about twice what is spent in Canada. If the United States were able to administer its healthcare system as efficiently as in Canada, there would be about $350 billion of annual healthcare savings — money that you could direct toward actual patient care — paying nurses, doctors, hospitals, drug companies, and equipment manufacturers.
People will say incorrectly that insurance companies consume 31% of total healthcare costs in their overhead, which is not true. They’ve missed the distinction between the part of administration that’s just insurance overhead and the part that is generated because of private insurance. Private insurance has its own overhead and then, because we have a multi-payer system, hospitals and doctors have very high overhead as well.
As a general rule, the administrative costs are much lower if you have a single payer. And if you have private insurance that, of course, implies multiple payers, and that always increases your administrative costs.
Q: Doctors at 26.9% stood out. What’s going on there?
There’s a fairly complicated methodology used to get to that number. We had to go through census data, which allowed us to identify the types of workers in doctors’ offices in the U.S. and Canada. Also there were surveys in both countries about how much time doctors and staff spent on administrative issues. So it was a combination of census bureau data and survey data.
In a doctor’s office in the United States you generally have several employees who spend most or all of their time on payment-related issues. For example, consider the reception staff; patient scheduling is a clinical task, but receptionists also spend a lot of time taking insurance information, getting prior approvals, and handling billing slips. Often there’s also a billing person who takes the material from the reception staff and the doctors and processes it again. A typical doctor’s office spends a lot of the personnel budget on billing-related activities.
That’s in contrast to Canada where billing is really easy. All your bills go to the single payer. It’s simplified so that a doctor could function without office staff at all. You probably need someone to make your appointments, but other than that you wouldn’t necessarily need office staff.
Q: What do you think would be a reasonable number for administrative costs overall?
Well, Canada does 16.7%, so I think that would be reasonable. They do have a little bit of private insurance, but it’s a small enough sector that it doesn’t contribute much to overhead. That 16.7% is not theoretical; it’s what a country very similar to the United States is actually doing.
Q: Since your latest data in ’99, what do you think the trend has been?
I guarantee it hasn’t gotten better. Either it’s stayed the same or it’s gotten worse.
Interview conducted and edited by Ted O’Callahan
Source: Woolhandler S, Campbell T, Himmelstein DU, "Costs of health care administration in the United States and Canada," New England Journal of Medicine, 349: 8, (2003): 768-775.