Shared Decision Making for Hip and Knee Replacement Candidates

Osteoarthritis (a.k.a "degenerative arthritis) of the hip and knee just... sucks. Characterized by activity-related pain in the affected joint, many otherwise physically fit persons have to resort to pills, injections and, finally, an appointment with an orthopedic surgeon to talk about joint replacement surgery.

What is less appreciated is that osteoarthritis can have a waxing and waning course with periods of relative remissions. What's more, conservative treatment options can lessen or delay the need for surgery. Last but not least, the surgery itself involves months of recovery and the possibility of a nasty complication.

The primary care physician Disease Management Care Blog presided over this many times with its arthritis patients.  It was generally reluctant to refer a patient to an orthopedic surgeon because it knew that the patients would be more interested in the potential benefits and pay less attention to the downsides of surgery.

Enter shared decision making (SDM). Defined as care that is respectful of and responsive to individual patient preferences, needs, and values and ensures that patient values guide all clinical decisions, the premise is that by giving patients the information they need, they'll be able to ultimately determine the course of their care.  That would include patients with severe hip or knee osteoarthritis who are thinking about surgery but who also need to consider the option of conservative management.

That's why this just-published Health Affairs study is noteworthy. All the 27 orthopedic surgeons in the 5 Group Health Cooperative clinics introduced shared decision making (SDM) for patients who were being evaluated with knee or hip osteoarthritis.  The intervention consisted of DVDs and booklets (from this company) that were ordered by the surgeon prior to an appointment.  The materials could also be viewed on Group Health's website at any time.

The study itself was quasi-experimental.  To be included in the study, patients had to 1) have knee or hip arthritis, 2) ) be continuously enrolled in the Group Health Plan for 12 months prior to the orthopedic clinic visit and 3) have a visit itself that was first index visit by the patient for that problem being evaluated by that particular specialty.

Outcomes from the 18 months of the SDM intervention period (January 2009 through July of 2010) were compared to the observation period of January 2007 through July of 2008.

Recall that the surgeon had to proactively order the SDM prior to the visit.  As a result, only 41% of the hip patients and 28% of the knee patients received the DVD, pamphlet or viewed the on-line materials.
 
Nonetheless, during the 6 months after the initial visit, the SDM patient population had 0.34 hip operations per 180 person-days (your DMCB offers an explanation of this counter-intuitive metric below*), compared to the control population of 0.46.  The difference was statistically significant. 

There was also a statistically significant reduction in knee operations: 0.09 per 180 person-days vs 0.16 per 180 person-days. 

All the differences held up after the authors statistically adjusted for differences in age, sex, obesity, co-morbid conditions, use of prior x-rays, joint injections, insurance factors and the clinic site.

Like all good authors writing in a high quality journal, they point out that this research was not pristine. The comparison period may not have been a representative baseline and, from 2008 to 2009, other factors may have caused a drop in hip and knee surgeries.

Nonetheless, this is an example of a "real world" study that credibly demonstrates that when osteoarthritis patients are exposed to SDM, more will opt for conservative management.  While that helps decrease health care utilization and ultimately costs, that's not the most important point: the patients who really wanted surgery got it and the patients who were less sure about the benefits of surgery chose not to have it.  What's more, this didn't involve a lot of expensive face-to-face care management, it involved some DVDs.

The DMCB cautions that this successful study was carried out in a highly integrated delivery system and may not be transferable to other practice settings.  That being said, as Accountable Care Organizations struggle to meet their patients' expectations and save money, this application of SDM may represent an important option.

*The DMCB interprets "180 patient days" as one patient being followed for the entire 6 months of the study.  If that's correct, the average SDM knee patient referred to a Group Health orthopedist had a 34% chance of getting surgery versus a 46% chance in the prior control group.  For the knee patients, it was 9% vs. 16%

Osteoarthitis Disease Management: One of the "Big Six?"

The Disease Management Care Blog has always wondered why there aren't more disease management programs for osteoarthritis (OA). Sure, some population health vendors offer them (for example) but OA never achieved status as one of the "Big Five" (mentioned on page 5 here). That's curious, because the condition is quite prevalent, self-care has a role in limiting disability and big pricey surgery awaits at the end of the line. There is no shortage of rigorous peer-reviewed studies that show that patient education programs can reduce pain, increase mobility and reduce physician visits (for example, here and here).

By the way, osteoarthritis shouldn't be confused with other types of joint inflammation, such as "rheumatoid" arthritis. The former is primarily a disease of middle aged and older individuals involving the cartilage of the major joints (back, hip, knees and hands) typically accompanied by bony thickening. It is often a diagnosis of exclusion, made after other causes of joint inflammation are ruled out. Criteria on making the diagnosis can be found here.

Thanks to a randomized prospective study appearing in the latest Annals of Internal Medicine, it may be time to reconsider whether OA should become one of the "Big Six."Researchers at Durham VA scoured their electronic health record for patients with the diagnosis of OA, mailed them letters about the study and then telephoned them. Out of 3477 persons with an EHR diagnosis, 523 were successfully recruited for participation while 461 completed the study.

The patients were randomly allocated to one of three treatment arms:

1) the intervention OA self-management group (N=172) which received "grounded in social cognitive theory," education, goals setting, action plans with monthly phone calls from an educator,

2) the control health education group (N=172), which received generic education on a host of health issues plus phone monthly calls from an educator, and

3) usual care (N=171).

At the end of 12 months, measures using standardized self-assessment scales showed the intervention group was experiencing less pain and also had better mobility and flexibility. The intervention itself was estimated to cost a total of about $118 per participant versus $63 in the health education control group. 172 patients times $118 is just over $20,000. While avoided surgeries were not measured in this study, the figure of $20,000 contrasts with the cost of a single total knee replacement.

The DMCB was also struck by the similarity of the research protocol to standard disease management vendor approaches to covered populations. Instead of using an EHR to find patients, vendors use insurance claims. Just like the Durham Va researchers, they also send letters and then telephone patients with a less than 100% success rate. In addition, the economic game of cost "savings" or "ROI" success or failure can hinge on one or two patients. The only difference is that the authors never used the terms "disease management" or "population health management."

These study results, combined with more than a decade of other research, make the DMCB suspect that an $118 per member per year (PMPY) program like this would ultimately yield significant savings for a managed care health insurance or self insured sponsor. While it could benefit all patients with OA, the DMCB wonders if it couldn't be targeted at patients that have been referred by primary care physicians for specialty care, since those patients are probably at the greatest risk for having an elective total knee replacement.

Last but not least, note that evidence-based guidelines on the treatment of OA strongly emphasize the role of patient education. It may be time to update them so that they emphasize the role of cognitive psychology, goal setting, action plans and remote coaching. Of course, the guidelines could sum up all that and more with two simple words: "disease management."

Image from Wikipedia. Because the translucent cartilage that lines the joint is diseased, the bones that make up the knee are not separated and have begun to touch and rub on each other. There is also lipping and a knobby appearance, thought to occur as a result of the joint inflammation leading to attempts to heal.