Why It's Difficult to Calculate Provider-Level Cesarean Rates
Wednesday, January 18, 2012 at 6:45AMGuest post by Jill’s Friend, MD
I was asked my cesarean rate recently and gave my stock answer of “I do not know. I have never calculated the percentage because it is as low as possible, so why bother with the number?”
This.
Was.
The.
Wrong.
Answer.
I then took the time to calculate my section rate [Editor’s note: Rates were calculated under extreme duress]. I picked 2010 because it was handy. The 2011 calendar year would have been harder as all the data might not be in. Easy pickings right? Punch up the server, put in the section and vag codes and voila. WRONG! I noted that there were weird spikes in the cut-to-cure rates in certain months. Breaking it down further, I noted that the rate more than doubled for my patients on the weekends when I was not on call. OK, I needed to look at just the deliveries I did so I went through the call schedule and excluded all deliveries done on these weekends. This took over an hour.
Next, I needed to exclude deliveries by my non-VBACing partner. There is no way to separate these out except to look at every delivery she did, which was easier than going through all of mine. These deliveries excluded, and another hour lost, I now had the total deliveries I had performed myself on my patients. As deliveries performed for other docs were billed by them they did not show up in my server. Unreferred would have been trickier, but since they are my billing, I left them in despite a near 50% section rate for these patients. I covered multiple weekends for a small hospital and rural health care clinic. I called and got all of my deliveries from them even though I did not bill for them. Trying to break out primary from repeat cesareans was difficult until I found the code for scar revision. I clean up the old scar nearly 100% of the time so it made the repeats stand out. The results are a 22% section rate with with a 10.2% primary rate. FYI, I had 36 successful VBACs.
Now enlightened, I can see the problem with providers trying to be accurate about their rates. It is hard to pin down the exact number because the perfect criteria to determine if it was “yours” is elusive. My rate is probably higher because some of the sections on weekends were scheduled sections that presented in labor. Others were complications notorious for occurring after 5 p.m. In a larger group the rates would be even harder to individualize but most members of a clan will follow the same hunter/gatherer patterns; therefore, the group percentage as a whole should be easy to render. Hospital rates likewise should not be too misleading as they tend to have like practitioners even amongst different groups. One group among four or five usually will not produce too much of an anomaly.
So how do you select the person or group that will give you the best-standard-of-care shot at a normal delivery? Beats the hell out out me. There is no place to adequately look up self-reported rates. There is no way to verify these rates. There is no policing of these rates. Jesus, I cannot even confirm my own!
I do not personally know of any practitioner or group that posts their numbers and the hospitals have no clue who does what for whom. They rarely seem to even get the numbers close. All I can say is good luck in your search and let me know when the veil on true section rates by provider is lifted. It will either make me laugh my arse off or throw up.
Jill’s Friend, MD, has a cesarean rate that is half of the local and state rate and works way too much (which is why Dr. Jill’s Friend has a high VBAC rate and low primary cesarean rate). Jill’s Friend, MD, is going to kill her for defiling a perfectly good post with the Philosoraptor meme.
Reader Comments (17)
Question for the author: is the right higher or lower than you expected? Or maybe just about right? And kudos on beating the pants off of your local colleagues.
I can answer that. It's a tad lower than he expected. I am sure he would be showering himself with confetti but Wednesday is surgery day and there's that whole sterile field issue that precludes a proper celebration. He doesn't need a lawsuit for a retained party hat.
Thanks for sharing. I have heard that cesarean section rates are misleading because of all of the factors that went into this calculation.
Thanks for posting this. It does help highlight some of the nuances in reporting. That said, I'm not sure why births can't simply be broken down into "vaginal" and "cesarean" and then have sub-categories in those - assisted vaginal (vacuum, forceps), unassisted vaginal/ planned cesarean, unplanned cesarean, emergency cesarean with all of the reasons given for each. Each reason could then be listed as a sub-category so that you could see the patterns (are there a lot of births moved to cesarean because the baby is "too big" or the labor "stalled"?).
I know going back through data would be cumbersome, but once the system was set up, it would be easy to add births under their category as they happened. It may seem like a lot of work, but could definitely show some interesting patterns that could lead to changes in behavior or strategies to continue on with good practice (as in the case of this practitioner). I know it sounds like a lot, but also know that medical professionals are great with math, science, and research so this wouldn't be bending their skill set. More than even for reporting purposes, organizing birth outcomes could provide great insight to the providers so they can see what needs to change and what is going well. That's a personal accountability that could be welcomed.
"could provide great insight to the providers so they can see what needs to change and what is going well"
Unless you don't really care to know.
I'll expand on that. Saying that there is a need for improvement is the same as saying that there is something wrong with the way things are being done. There are some people that have a more difficult time handing criticism than others. Google the work of ethicist John Banja. That's all I am going to say about that. That, and that this issue requires a higher level of sensitivity than I anticipated.
I think it is great and admirable that you put so much time into answering this question.
However, I think for the average woman who would ask about a particular care provider's cesarean rate, what she really wants to know is "if I choose this practice for my care, what are my odds of having a cesarean?"
Given that...excluding "your" clients who delivered with other doctors on weekends would not get that answer, neither would including the patients from other practices that you delivered get that answer. Her question is not really about you, it is about her.
I find it odd that in the late 90's when I was pregnant with my first two children I had no problem getting an answer to this question from my OB practices, but now it is so difficult to get.
In my research on cesarean reduction efforts, I've found that it is possible for hospitals to provide regular updates for clinicians about their rates and to contextualize them anonymously within their peer group at that facility. And regular evaluation and collegial data-sharing does impact rates. It is absolutely reasonable to assume that many doctors just don't know how high their rates are or how their performances compares to their peers. It isn't practical for the individual clinicians to do this, though, it should be supported on an institutional level.
Here are the replies I got:
1. Regarding how it was no problem in the late 90's to get disclosure of personal/group data... he says they probably just made it up.
2. "It isn't practical for the individual clinicians to do this, though, it should be supported on an institutional level." - That was something we talked about, too. It's unrealistic to think that someone is going to sit down and do all of this, honestly (without someone double dog-daring you to do it).
There was a midwifery practice in NYC that used to post their cesarean rates on their website, but they don't seem to any more... it looks like the practice has changed and they've gotten a new website too. I can't think of anyone else I've ever seen who had their rates out for the world to see!
Jill, I double dog dare you to have your MD friend find out how Birthrisk.com has already solved his problem. Data entry takes less than two minutes per birth and can be accomplished by existing hospital or office personnel.
The most important piece of information to keep in mind when discussing cesarean birth rates is that the risk that labor will result in a cesarean birth is affected by BOTH a woman’s physical characteristics AND the labor management strategies used by the obstetrical care provider. Give me a population of twenty year olds having their second vaginal birth and my total cesarean birth rate will be well below 5%, however if my patients are all in their upper thirties and having their first birth, a 30% cesarean birth rate may be exceptional.
My Birthrisk Cesarean Birth Measure is the only measure available that takes into account the physical characteristics of both mother and baby so that we can finally measure the effects of how different obstetrical care providers manage labor. Once all obstetrical care providers obtain their Birthrisk Cesarean Birth Measure then women will finally be able to compare apples to apples. Any obstetrical care provider that wants to obtain their measure can do so for free. Any woman who is having a baby and wants to know how women with similar physical characteristics fared during labor can obtain a free Cesarean Rate History® at www.Birthrisk.com.
"I noted that the rate more than doubled for my patients on the weekends when I was not on call."
This is pretty frightening, though I suppose that since your rate is so low (1/2 the average) that it just bumps their odds up to "average" if you're not on call that particular weekend :(
Amazing that an OB would highjack this post to plug HIS website. This shaman appears to quantify your section risk with his own data. Nothing in your comment, Dr Roman, has anything to do with ascertaining the post delivery true cesarean rate of my practice. Please take out an ad in the green journal and post how many practitioners "register" as members for your obvious breakthrough calculations. Meanwhile, stop plugging your crap in my post.
Err, I mean Jill's post of my letter.
Readers, here is an astonishing non-published bit of data. In an attempt to find the highest associated risk factor for a cesarean section a junior resident of mine researched his butt off going through reams of data points dating back five years. This was collected in a public hospital with no infertility guys around (no quints, etc). The result, and reason not to publish, was (wait for it......) a hyphenated last name. Yes, this single point emerged with a near 100% rate. Still think we can accurately predict?????
"a hyphenated last name" - holy crap, that explains everything (says the woman who not only hyphenated her own last name, but whose husband and children are all hyphenated as well.) I never stood a chance. Luckily the hyphenation risk factor must only be for primary births - since my second two came out the door, not the window. ;)
Dear MD,
My data represents over half a million births that were attended by 2,915 different obstetrical care providers working out of 109 different hospitals. However, you are correct that nothing in my comment ascertains the true cesarean rate of your practice. My work is a quantum leap ahead of using the cesarean “rate” as a meaningful measure. ACOG recognized this same problem over 12 years ago and the Joint Commission is using the cesarean birth “measure” created by Dr. Main and his group in California as a result of this.
It seems to me that your attack on my comment offering you a solution to your problem would indicate that you really don’t want a solution. Other than your insults, you have not provided any meaningful evaluation of my work. There is a 38 minute PowerPoint presentation on my website that explains my work in detail. Please take the time to view that presentation before you refer to my work as “crap”.
I apologize. I was not aware I had a problem. Please do tell how your perinatal cesarean risk prediction calculator helps with my postpartum cesarean rate calculation. I, in my backward southern state of stupidity, can see no connection between the two. Alas, I talk slow and must therefore be an idiot. I bow to your universally applicable, ACOG endorsed means of providing me, the lowly non-academic Ob, a means of abandoning any trial of labor in favor of just whisking the high riskers off to the cesarean assembly line. I'm just not ed-u-kate-ed enough to understand.
However, after 22 years of practice with now pushing ten thousand deliveries I do feel I am starting to get the hang of this job. Now, which Green Journal are you published in? I seriously would like to read the peer reviewed version with comments about your calculator.
Dear MD,
I am not sure why you refer to my work as a “perinatal cesarean risk prediction calculator” when that is not what my software does. My Cesarean Rate History® software tool provides the outcome of the last 100 similar women who have labored. This provides information not a prediction. I agree with you that 22 years of practice and ten thousand deliveries provides a wealth of experience. I am not sure why adding a few hundred thousand deliveries to your experience would not be a good thing. Imagine having the experience of all those deliveries as a first year resident. It is for this reason that my software is provided at no cost for resident training.
My Cesarean Rate Tracker® software tool providers a central location where an obstetrical care provider can store the results of their deliveries and compare the outcome of their deliveries to other providers in a confidential manner. This would help you with your postpartum cesarean rate calculation.
Your letter states; “So how do you select the person or group that will give you the best-standard-of-care shot at a normal delivery? Beats the hell out out me.” I believe that the answer to this question is that we need to look at a cesarean birth “measure” and not a “rate”. A cesarean birth measure takes into account the risk factors of an obstetrical care provider’s patient population therefore; it can be used to measure the results of an obstetrical care provider’s labor management strategies.
The Green Journal has reviewed my work and chose not to publish it “on the basis of priority”. The reviewers had many positive comments and commended me on my work. I would be happy to provide you with a copy of that research paper.