What is practice variation in obstetrics and why should I care?
By Adriana Arcia, PhD, RN
When it comes to our health, we all want the very best care. But what is "best"? For any given health-related scenario, two healthcare providers may come to very different conclusions (example: "cut them out" vs. "wait and see") about what constitutes the “best” treatment decision. The resulting difference in rates of treatment is practice variation.
A diversity of opinion is to be expected when we don’t have enough scientific evidence to say definitively that one treatment is better or worse than another such as with the most cutting edge cancer treatments. However, for most common conditions, there are guidelines and standards of care that help healthcare providers make treatment decisions that align with the best scientific evidence. According to a 2001 report from the Institute of Medicine (IOM), "care should not vary illogically from clinician to clinician or from place to place" (p. 4). Put another way, if care varies, there had better be a good reason.
So, why should you care about practice variation? Because, "…one of the most fundamental principles in quality assessment and control is that unwarranted variation in a product or process generally equates to poor quality. Conversely, as quality improves, variation will diminish" (Clark, Belfort, Hankins, Meyers, & Houser, 2007, p. 526.e1). In other words, wide variations in care are problematic because they mean that some practitioners are making decisions that deviate from accepted best practices and therefore their patients are not getting the best quality care. Individuality of style may be good in fashion but when it comes to healthcare, consistency is a virtue.
In obstetrics, practice variation is most readily observable in the rates at which healthcare providers (typically, obstetricians and midwives) employ/recommend interventions such as induction of labor, epidural, and cesarean section. If you’re a regular cesareanrates.com reader, you are most likely already aware that cesarean rates vary, sometimes dramatically so, between facilities. For example, the cesarean rates of two facilities in West Virginia vary by 54 percentage points despite being only 60 miles apart. Although not usually that stark, many such examples of variation can be found around the country; poke around the site and see. [For a good scholarly analysis, check out the work of Baicker, Buckles, and Chandra (2006).]
Of course, to compare one facility to another using only their overall cesarean rates is to use a rather blunt instrument; practice variation is more complicated than that. Overall cesarean rates don’t account for the fact that some hospitals specialize in treating high-risk women whose clinical needs are complex. Nor do they account for practice variation between individual providers. In one scientific study, 11 obstetricians (in one community hospital serving women with uniformly low obstetric risk and with no differences in neonatal outcomes) had primary cesarean rates that ranged from 9.6% to 31.8% (Goyert, Bottoms, Treadwell, & Nehra, 1989). The study is over 20 years old (here’s another one like it), but the example is instructive.
A more recent example of an apples-to-apples comparison of practice variation is even more intriguing. Janssen and colleagues (2009) compared 2,899 women who intended to have a planned home birth with a midwife to a group of women attended in hospital by the same midwives and a group attended in hospital by physicians. All the women in the study were sufficiently low risk to be good candidates for home birth. There were no maternal deaths and the (very low) neonatal mortality rate was not statistically different between groups. Given that women who elect planned home birth are very motivated to have low-intervention deliveries, it’s not terribly surprising that the home birth group had the lowest rates of cesarean, epidural, etc. When it came to interventions that are sensitive to clinician judgment (e.g., augmentation of labor, assisted vaginal delivery, episiotomy, etc.), physicians had the highest rates all around. What was surprising (to me, at least) is that the same group of midwives practiced differently depending on the setting. For instance, they performed episiotomies twice as frequently (3.1% vs. 6.8%) in the hospital as in the home. The point of this example is that practice variation is evident not just between states, or facilities, or even individual providers; clinicians’ practice may vary even according to their practice environment.
At this point, you may be asking yourself, "WHY?!?" That’s the $5 billion per year question. My opinion is that practice variation is due to a complex interplay of local practice culture and clinicians’ personal inclinations served with a small side of financial incentives, convenience, and defensive medicine. [For a more thorough, less glib take on this topic, see pp. 22-31 of my dissertation.]
So, let’s circle back to the beginning and revisit the question of what constitutes the "best" in obstetric care. To say that there is disagreement on this point is an understatement. I think most clinicians can agree that practice variation is an indicator of inconsistent quality of care and therefore not the best we have to offer. In obstetrics, the variation is so wide that Clark and colleagues (2007) characterized it as "disturbing" (p. 526.e2) and suggestive of “a pattern of almost random decision making” (p. 526.e1, emphasis mine). If that last quote sends a cold chill down your spine as it does to me, you’ll concur that having access to data about practice patterns is vitally important. The IOM doesn’t know it, but they agree with me:
"Transparency is necessary. The [healthcare] system should make available to patients and their families information that enables them to make informed decisions when selecting health plan, hospital, or clinical practice, or when choosing among alternative treatments. This should include information describing the system’s performance on safety, evidence-based practice, and patient satisfaction" (p. 4).
Transparency is not just beneficial to patients. Data about clinical practice patterns can help healthcare providers evaluate the quality of their care using objective metrics. Sadly, we have a lot informational infrastructure to build before we can achieve the transparency called for by the IOM.
That is why cesareanrates.com is such an invaluable resource. I have to commend Jill Arnold for her indefatigable efforts in tracking down birth statistics, crunching the numbers, and making them public. Evaluating a facility (or provider) solely by their cesarean rate may be inadvisable, but using that metric as part of a broader information-gathering strategy is far better than simply picking off of the top of the list provided by your insurance company. Being an actively informed participant is often the best way to increase the likelihood that you will get safe, high-quality care.
Adriana Arcia, PhD, RN is a researcher, registered professional nurse, and doula. As a researcher, she wants you to make data-driven decisions. As a nurse, she wants you to cover your cough and wash your hands. As a doula, she wants you to listen to and trust your body.
Arcia, A. (2011). Predictors of Nulliparas’ Childbirth Preferences (Doctoral Dissertation). Retrieved from http://scholarlyrepository.miami.edu/oa_dissertations/671/
Baicker, K., Buckles, K. S., Chandra, A. (2006). Geographic variation in the appropriate use of cesarean delivery. Health Affairs, 25, w355-w367. Retrieved from http://content.healthaffairs.org/content/25/5/w355.full.pdf
Clark, S. L., Belfort, M. A., Hankins, G. D. V., Meyers, J. A., Houser, F. M. (2007). Variation in the rates of operative delivery in the United States. American Journal of Obstetrics & Gynecology. 196, 526.e1-526.e5. http://download.journals.elsevierhealth.com/pdfs/journals/0002-9378/PIIS0002937807000956.main-abr.pdf?jid=ymob
DeMott, R. K., & Sandmire, H. F. (1990). The Green Bay cesarean section study. I. The physician factor as a determinant of cesarean birth rates. American Journal of Obstetrics & Gynecology, 162(6), 1539-1599. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2360593
Goyert, G. L., Bottoms, S. F., Treadwell, M. C., & Nehra, P. C. (1989). The physician factor in cesarean birth rates. New England Journal of Medicine, 320(11), 706-709. Retrieved from http://www.nejm.org/doi/full/10.1056/NEJM198903163201106
Institute of Medicine (2001). Crossing the Quality Chasm: A New Health System for the 21st Century (Report in brief). Washington, D.C: National Academy Press. Retrieved from http://www.iom.edu/~/media/Files/Report%20Files/2001/Crossing-the-Quality-Chasm/Quality%20Chasm%202001%20%20report%20brief.pdf
Janssen, P. A., Saxell, L., Page, L. A., Klein, M. C., Liston, R. M., & Lee, S. K. (2009). Outcomes of planned home birth with registered midwife versus planned hospital birth with midwife or physician. Canadian Medical Association Journal, 181(6-7), 377-383. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2742137/pdf/1810377.pdf
Schytt, E. & Waldenstrom, U. (2010). Epidural analgesia for labor pain: Whose choice? Acta Obstetricia et Gyencologica, 89, 238-242. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/19824867