Physician Administrative Costs in the US vs. Canada

The US spends far more per capita on healthcare than any other developed country -- $7,538 per person, compared to $3,129 in the UK, $4,079 in Canada, and $5,003 in Norway (the second-biggest spender), according to 2008 totals compiled by the Kaiser Family Foundation. One contributor to our high healthcare costs is high administrative costs, which is the natural consequence of having hundreds of different insurance plans with different policies, networks, and rates. A new study in the journal Health Affairs focuses on one aspect of administrative costs: the time physician practices spend interacting with payers. They surveyed US and Ontario practices and quantified just how much time and money the US proliferation of payers costs physicians.

Canada has a single-payer system, while the US has a wide variety of insurance plans (as well as a large uninsured population). In theory, the proliferation of insurance plans fosters competition that can improve price and quality for plan members as the plans adjust benefit design and cost-sharing requirements. To control members' healthcare costs, plans often do the following kinds of things:

• Establish networks of physicians and negotiate payment rates with them
• Limit the number of visits covered in a calendar year (e.g., one physical per year, 40 therapy sessions, etc.)
• Require varying co-payments and/or co-insurance from patients, depending which providers they visit
• Establish drug formularies and/or tiered pricing structures that encourage the prescribing and purchasing of certain drugs and not others
• Limit coverage to visits and services based on the patient's diagnosis
• Require prior authorization for certain services and procedures

In some cases, these policies can actually contribute to the quality of patient care. Research has found that providers don't always deliver care that follows clinical guidelines; for instance, a 2003 study by Elizabeth McGlynn et al examined medical records from consenting participants in 12 metropolitan areas and found that 11.3% of the participants "received care that was not recommended and was potentially harmful." If your physician is prescribing a drug or procedure that's not appropriate for your condition and has the potential to harm you, it might not be such a bad thing for someone to be denying coverage for it. Sometimes, of course, an insurer will deny something that might seem inappropriate to a non-expert with limited information, and the provider will then have to spend time going back and forth with the insurer to explain why it is, in fact, appropriate. (And getting these things resolved may take weeks of frustration.)

Insurance plans can also encourage the use of high-value care and discourage the use of low-value care through what's called value-based insurance design -- basically, varying cost-sharing requirements based on how much a particular kind of care is likely to contribute to your health. For instance, a plan could reduce co-payments for hypertension or diabetes medication to encourage patients to control these conditions, and charge higher co-payments for MRIs ordered within the first six weeks after the onset of back pain. (Many cases of back pain resolve within six weeks, so MRIs aren't recommended until the pain has lasted beyond that time frame, unless there are other problematic symptoms like problems with urination.)

Interactions between physicians' offices and payers
In exchange for the potential cost-containment and quality benefits of having many different insurance plans, patients and providers pay for the complexity inherent in such a system. Many physician practices pay multiple staff people to do nothing but bill insurance companies and go back and forth about denials and authorizations. When providers want to prescribe a drug, give a referral, or order a test or procedure, they have to think about what the patient's insurance will cover and whether all the necessary requirements have been met. For instance, some insurers will only cover a branded drug or experimental treatment if a patient has tried other generic or standard therapies without success. If all the paperwork isn't filled out correctly, claims will be denied and someone from the doctor's office -- as well as the patient, in many cases -- will have to spend more time going back and forth with the insurer about it.

Back in 2009, Lawrence Casalino and colleagues published a study in Health Affairs that quantified the cost -- in both hours and dollars -- of physician practices' interactions with insurance plans. They surveyed physicians and administrators from physician practices of different sizes (excluding HMOs, hospital physicians, and those, like plastic surgeons, with many self-pay patients) about the minutes per typical day the practice's physicians, nursing staff, and clerical staff spent on specific categories of interaction with health plans: authorization, formulary, claims/billing, credentialing, contracting, and quality data. They excluded extreme outliers from their analysis, and to ensure that the survey was not capturing time spent interacting with Medicare and Medicaid they both asked respondents to report only time spent interacting with health plans and reduced the claims/billing time estimates by 38% (the percent of charges attributable to those payers).

Their analysis of 895 completed surveys found the following average time spent interacting with health plans:

• Physicians: 43 minutes per day, or 3 hours per week (median: 28 minutes per day, 1.9 hours per week)
• RNs, MAs, LPNs: 3.8 minutes per practice physician per day, or 19.1 hours per physician per week (median: 1.8 hours per day, 9.1 hours per week)
• Clerical staff: 7.2 hours per physician per day, or 35.9 hours per week (median: 5.9 hours per day, 29.8 hours per week)

The large differences between the mean and median values suggests that several (fewer than half) of the respondents reported very large time expenditures, which increased the overall average but not the median -- and this was after extreme outliers had been trimmed. The authors note that average time expenditures were higher for primary-care physicians than for specialists, and physicians in solo or two-person practices spent significantly more time interacting with health plans than did their counterparts in practices with ten or more physicians.

The authors converted that time into dollars and arrived at an average of $68,274 per physician per year interacting with health plans (median value: $51,043). They estimate that this works out to $31 billion each year spent on physician practices' interactions with health plans ($23.2 billion if you use the median value rather than the mean). This doesn't include other overhead costs (phone, fax, etc.) related to the interactions. More importantly, it doesn't include the health-plan-interaction costs of hospitals, which account for around 30% of US healthcare expenditures.

Fast forward to 2011. Casolino and a slightly different group of colleagues (Dante Morra et al) have just published another study in Health Affairs. They surveyed office-based physicians and administrators in Ontario (again, excluding hospital physicians and those with large numbers of self-pay patients) about how much time they spend interacting with payers. For comparability with the US results, they "report adjusted Canadian costs as if the Canadian physicians and staff were paid at US rates." They also created new estimates of US physician practices to include time spent interacting with Medicare and Medicaid, which were excluded from the 2009 results. Here are there results (only means were reported):

• Canadian physicians: 2.2 hours per week
• US physicians: 3.4 hours per week
• Canadian nursing staff: 2.5 hours per week
• US nursing staff: 20.6 hours per week
• Canadian clerical staff: 15.9 hours per week
• US clerical staff: 53.1 hours per week

When the researchers adjusted Ontario costs to US salary rates, they concluded that Ontario practices spend $21,335 per physician on interacting with payers, compared to $82,975 in the US.

In both articles, the authors point to some limitations, including the fact that the data come from surveys rather than from direct observation, and that the practices that chose to respond to the surveys might have been those who spent more time on payer interactions. It's possible that the gap between US and Canadian practices' time spent interacting with payers is smaller than what this study found -- but it's also possible that it's larger.

Are the actual and potential benefits of having many competing health insurance plans worth spending nearly four times as much Canada's single-payer system does on this one component of administrative costs alone?

Morra, D., Nicholson, S., Levinson, W., Gans, D., Hammons, T., & Casalino, L. (2011). US Physician Practices Versus Canadians: Spending Nearly Four Times As Much Money Interacting With Payers Health Affairs, 30 (8), 1443-1450 DOI: 10.1377/hlthaff.2010.0893

Casalino, L., Nicholson, S., Gans, D., Hammons, T., Morra, D., Karrison, T., & Levinson, W. (2009). What Does It Cost Physician Practices To Interact With Health Insurance Plans? Health Affairs, 28 (4) DOI: 10.1377/hlthaff.28.4.w533