A billion-dollar bust? Let's examine that claim closely

Just a few days after I published my last blog entry, lauding the new era of co-operation between doctors and government, Ruth Lavergne and Kim McGrail, researchers at the University of BC Centre for Health Services and Policy Research (CHSPR), published a commentary in the Globe and Mail that was highly critical of one of the key programs under these changes.

You can read their online commentary here, http://bit.ly/1yVovzH/ ,but the headline sums up their stance pretty succinctly:"Was BC's push for better primary care a $1-billion bust?" The article is based on their study, published in the May2014 Healthcare Policy, which examined 20 years of patterns of family doctors' fee-for-service billings in BC. They use this administrative data to come to the conclusion that patients have not been served by the program of incentive payments to GPs brought in under this joint doctors/government culture. You can read the abstract of their study here: http://www.longwoods.com/content/23782

"Isn't this at direct odds with your perspective, Anne?" I was asked in an email by one of my thoughtful blog readers.

It is, indeed. But I think their study has serious limitations, makes assumptions not backed up by their own or other research evidence, and comes to conclusions that cannot be made from their data. Moreover, they completely ignore patient health outcomes, which as I explained in the previous blog, was the unifying factor that got the perpetually-wrangling docs vs government to agree on a program in the first place. As I noted in the previous blog, I am working with these committees to explain their rationale and activities.

I am going to explain here why I think this study is weak and inconclusive. It is going to be a bit of a long entry, so bear with me. You can see whether or not you agree with me. I welcome debate and discussion.

First I will summarize what the authors did and what they say they found. They used administrative health data linking patients' anonymous personal health numbers to individual general practitioner's fee-for-service billing data, beginning from 1991/92 through to 2009/10.

They looked at what they called four main "conceptual" measures : access, coordination, continuity, and comprehensiveness. In essence that meant they looked at, from the billing data, where the care was provided (doctor's office, house call, hospital, nursing home) who it was provided by (one doctor or more than one over the year); what time of day it was provided (after hours or in office hours), and what was done (they looked at PAP screens, mental health, maternity care, glucose and lipid testing.)

They concluded that access, coordination, continuity had all continued to decline. Interestingly for comprehensiveness they found only the number of individual GPs doing maternity care had decreased during the study period, but the number of GPs doing mental health care, PAP screens, and preventive tests like glucose and lipid testing had actually increased — remember that, because I will come back to that at the end. They conclude in the Globe commentary that patients would be better served by putting BC doctors on salary.

Here are my problems with the study:

1)It studies trends from 1991 through to 2010. The decline of primary care was well underway for more than a decade when the joint General Practice Services Committee (GPSC) was formed in 2003. The bulk of the GPSC incentives and the Practice Support Program that teaches doctors and staff both clinical and administrative skills on how to use the incentives started late 2007 and didn't gain force until 2008/9. Some incentives came in after 2010. How can the CHSPR study make blanket statements about success or failure based on 2 or 3 years of 18 years of trends data?

2.)There is no control group, or no comparison to what happened with GPs in other jurisdictions over the same time period. BC's changes applied across the board to all doctors, anywhere, but it was completely voluntary. CHSPR, by using billing codes cannot distinguish who was "on the program" and who was not. As with all programs there were early and late adopters. All these are in the same sample. They cannot show what might have happened to their data if no changes had been applied or a different model was taken in BC, in light of a huge decline in family practice across all Western nations.

3.)They make the assumption that seeing more than one doctor is bad for care (without showing its impact on patient outcomes.) We do know that going to walk in clinics, for a single issue and that doctor not being able to access the patient's full medical record, is not as good care as having a doctor know the whole patient history and seeing the full record. But in BC now many doctors are working in group practices, such as my GP. That means when I make an appointment at her office, if she is not available, I may see one of her associates but they all have access to my health record, my Rx history, my pap screen record etc. In the last year I have seen four doctors, all in my family doctor's office, all of which coordinated around my care. The CHSPR study of billing number patterns cannot distinguish group practice from walk in clinics, when an increasing number of GPs this past decade are working in group practices.

4.)They equate house calls, hospital care, and after hours care with patients having higher quality care and being seen in doctors' offices, during office hours, with patients having poorer quality care. Where is the evidence that house calls, hospitals and after hours is better care and being seen in the office is lower quality? BC has been focusing on managing chronic diseases, planning care and avoiding crises and hospitalizations. This data finding could equally mean that fewer patients are having crises. A doctor that does a house call takes more than five times as long —- keeping other patients who need care waiting. It is very inefficient. Does it lead to better care? We cannot say. It is likely better not to need a house call at all. BC has also introduced in 2010 incentives for telephone calls from GPs to patients at their home, which greatly reduces the need for house calls, which this study also did not pick up. How can they conclude patients are worse off now from this finding of location of care? They cannot. There is one finding, however, that concerns me: A 20-year steady decline in GPs visiting their long time patients in nursing homes. If we want GPs to support patients to the end of life, how we stop this trend should be examined.

5) They say patients are not being seen after hours by doctors and equate that patients are therefore being less well served. They cannot make that assumption in BC because if a doctor's practice or group practice has decided that they will regularly have one or more nights a week where they provide service into the evening, as many have done in the last 10 years, there is no fee code difference for time. There only is a fee code difference if the doc is called out on an emergency after regular hours. BC's changes are all aimed, again, at avoiding the crises and emergencies, and going to better managed care. It is better for doctors and better for patients to avoid those after hours emergency calls.

6)They don't look at system outcomes over the time frame. CIHI data shows that BC is now leading the country on a number of key health indicators of system performance:

  • Ambulatory care sensitive conditions:  Since 2005 we have the lowest "ambulatory care sensitive conditions." This is a horribly obtuse term for acute and chronic diseases that with good primary care management outside of hospital (ambulatory care) leads to reduced or fewer hospitalizations. It is seen, world-wide, as a measure of an effective primary care system. So BC has the lowest hospitalization rates for diabetes, asthma, COPD, congestive heart failure, high blood pressure, vaccine-preventable conditions, pneumonia and gastroenteritis.  This is good.
  • Avoidable mortality: BC has the lowest avoidable mortality in Canada, meaning we have the fewest people dying before 75. This is seen as a measure of the general health of the population — in BC we have the fewest smokers, the most exercisers, and the lowest rate of obesity, which really has nothing to do with our health system — but avoidable mortality is also seen as a general measure of overall health system performance.
  • Avoidable mortality for treatable causes: BC has the lowest rate of death under age 75 once a disease is found. So if you are diagnosed with diabetes, or high blood pressure, BC's death rate is the lowest in Canada. This is an important marker of quality of health system performance. Could it be lower? Yup, particularly among the poor in BC. The wealthy in BC have among the best rates for this measure in the world. Our poor very much less so. But socio economic status (SES) has a huge impact on health outcomes. The next huge jump in health improvement is making sure the poor are as healthy as the wealthy. No one has cracked that yet, but systems like Scandinavia and Netherlands, which have less inequality in wealth, have less inequality in health outcomes. If you are interested in reading more about national health indicators, and the impact of SES on health , read the 2013 CIHI Health Indicators Report, available for download here: https://secure.cihi.ca/estore/productFamily.htm?locale=en&pf=PFC2195&lang=en
  • Per capita spending on health by province. BC has the second lowest per capita spending on health, second only to Quebec. The Fraser Institute always uses this to say that BC is under spending, but our health indicators do not show this. In Quebec, the health indicators (near the bottom half) would suggest they may be under spending. In BC our lower per capita rate, coupled with better national health indicators noted above, would suggest that we are spending less money because we are keeping people healthier and out of hospital — which is what we want. That means we have a more cost efficient system. Could it be better? Yes compared to other countries.
  • Per capita spending for physicians. Highest rates according to CIHI are in Alberta and Ontario ($986 and $942) and BC is in the lower half of the pack at $844. So this shows that what we have been spending on physicians this last decade is actually less than other provinces and not so out of line, despite the "$1-billion bust" headline. CHSPR's focus on $1 billion is taken of proportion to the overall costs. Health care is mind-blowingly expensive. Our provincial system is $17 billion a year, so $1 billion over 10 years actually works out (when you factor the various growth rates of expenditure) to about 2.6 % of the annual health budget. In BC the health system burns through $1.2 million ever hour, so in the time I wrote this blog close to $3million was spent in BC! Check out this CIHI report for national and provincial health expenditure data if you want more info:  https://secure.cihi.ca/free_products/NHEXTrendsReport_EN.pdf

7.) The authors suggest that Ontario's primary care reform has been more effective. Ontario underwent changes at the same time in BC, but that province now actually has three systems of primary care for doctors. In another blog post I am going to take a closer look at the pros and cons of Ontario, compared to what BC has done, because that is a huge and complex topic. I know that this blog post has been going on too long and, you dear reader, are getting tired.

8.) Comprehensiveness.  I told you to remember the finding that mental health care, Pap screen, lipid testing and glucose testing had all improved over the time frame of their study. This is a point the study completely dismisses, yet this is precisely what the incentives were geared towards. These are the kinds of actions that appear to have an impact on patient health outcomes.

9.) Maternity care: One last point, I promise. Fewer doctors doing maternity care is concerning, but that is a finding that is happening everywhere across North America and has been declining for decades. In BC, the maternity incentive HAS NOT focused on supporting solo doctors doing maternity care because that is not so good for patients and most doctors do not want to do this alone. Solo doctors can only do so many deliveries a year and we know the more you do the better you are. Rather, BC has focused more on providing incentives to support a group of doctors joining together in group practice to share maternity care, to which other GPs in the region refer their pregnant patients. You can read about the details of the "Maternity Network Initiative" here: http://www.gpscbc.ca/family-practice-incentive/maternity-network-initiative . Evidence shows this concentrated care is better for patient outcomes and better for doctors' lives, meaning more docs are willing to deliver under this kind of model. We still need to find more ways to provide good maternity coverage outside of population centres. That is a real challenge. But the CHSPR study, finding fewer individual doctors are doing maternity care, cannot jump to the conclusion the situation is worse for patients now than prior to the incentives.

In closing, BC is by no means perfect. We still have a ways to go. One huge task is to ensure that everyone who wants a family doctor has one. One very promising change since the BC program is that more new medical graduates are choosing family practice, up from the nadir of 23 % in 2002 before the changes took effect to 39% in 2014. More family doctors are in the BC system and that helps more patients find doctors. It is not solved, but believe me this is a focus of a whole bunch of activities.

I believe we have a much better chance of doing the changes we need in a culture of collaboration between doctors and government that asks the question "what is truly best for patient care?" We should never be afraid of looking at evaluations to see whether we are actually accomplishing that.

 I fear, however,  the damage that can be done to this fragile cooperative relationship with simplistic studies that claim patients aren't benefiting, that only doctors are, and that don't look at the big picture.

Thanks readers, for staying with me on this long post. Again, I welcome debate and discussion. I don't have the answers, but I do try to present things as clearly as I see them.