Scott Ellison
January 20, 2009
PearlDiver data reveals details to a hidden, dangerous problem
Walking on ice is a dangerous business. But the real danger isn’t always a fall; it’s the cracks you never see. According to new PearlDiver analyst research numbers, this is true of total hip arthroplasty (THR) as well. Fractures occurring during a total hip arthroplasty or shortly after continue to pose problems for patients in early recovery. Using the PearlDiver Patient Records Database we can see the extent of this problem as well as which patients are most susceptible to these types of fractures.
During the winter months, lakes freeze across the northern United States. Inevitably people venture out on to the ice for skating, hockey, fishing, and any number of other reasons. For the majority of these winter-time adventurers the worst case scenario is a slip and fall ever so often. And most people out on the ice know that when the ice is cracked, they are in an area that may be dangerous. But each year there are reports of people who are unaware that the ice is too thin—people not seeing the cracks—and finding themselves in serious trouble.
This may be an extreme analogy; however it illustrates the point that it is the cracks we don’t see that may become the most problematic. This is true of total hip arthroplasty as well. Either during the procedure itself or during early recovery, many patients suffer fractures that remain unseen until there is a serious problem.
This problem of early fractures was brought to the surface by Daniel J. Berry, M.D. of the Mayo Clinic during the Current Concepts in Joint Replacement Winter 2007 meeting, , and more recently at the Winter 2008 meeting by Michael E. Berend, M.D. of the Center for Hip & Knee Surgery, who presented Peri-Operative Fractures: See It Before It Sees You! Drs. Berry and Berend’s studies are now confirmed and further substantiated by a review of the PearlDiver database of 136 million patient records
From 2004 through 2007, using a sample set of 21,000 patients under the age of 65 derived from the PearlDiver database, we found that 3.9% of all total hip arthroplasty patients were diagnosed with a femoral fracture within 180 days of their total hip surgery. The exact number was 841 patients diagnosed with a fracture complications associated with their arthroplasty surgery.
However, of the hip arthroplasty patients diagnosed with a fracture within the first six months following surgery, fully 48% were diagnosed with a fracture, and this was a surprise to us when we looked at the data, within the first five days following the procedure!
Looking at the scope of these numbers on a national level, according to the Healthcare Cost and Utilization Project provided by the United States Department of Health & Human Services, 104,576 patients under the age of 65 received total hip arthroplasties in 2006. From the information provided by the PearlDiver database as shown in Table 1, 4.1% of U.S. patients were diagnosed with a fracture within 180 days, accounting for approximately 4,288 patients under the age of 65 nationally.
Table 1: Percentage of Patients Diagnosed with Fracture
|
2004 |
2005 |
2006 |
2007 |
Total |
Within 2 Days |
0.9% |
0.8% |
1.0% |
1.0% |
0.9% |
Male |
0.6% |
0.5% |
0.7% |
0.7% |
0.6% |
Female |
1.2% |
1.2% |
1.3% |
1.3% |
1.3% |
Within 5 Days |
2.0% |
1.5% |
2.1% |
1.9% |
1.9% |
Male |
1.5% |
1.1% |
1.8% |
1.6% |
1.5% |
Female |
2.6% |
2.0% |
2.5% |
2.2% |
2.3% |
Within 30 Days |
3.5% |
3.0% |
3.3% |
3.0% |
3.1% |
Male |
3.1% |
2.3% |
2.5% |
2.4% |
2.5% |
Female |
3.9% |
3.8% |
4.1% |
3.6% |
3.8% |
Within 180 Days |
4.2% |
3.9% |
4.1% |
3.5% |
3.9% |
Male |
3.8% |
3.2% |
3.3% |
2.9% |
3.3% |
Female |
4.6% |
4.8% |
5.0% |
4.3% |
4.7% |
Source: PearlDiver Patient Records Database, 2004-2007
While this is a large group, the bigger question remains, how many of these patients left the procedure with a fracture or received one very shortly after experiencing surgery? This is where the numbers get really interesting. Of the patients within the PearlDiver database who were diagnosed with a fracture during the first six months of their recovery, 25.9% of the fractures diagnosed occurred within two days of having their total hip arthroplasty.
Sliding the number of days out slightly further, by five days after surgery another 25.8% of the patients were diagnosed with a fracture. Combining these time periods, of all patients who would suffer a fracture during the first six months of recovery 52% had been diagnosed with a fracture within five days post procedure!
Certainly the first several months of recovery from a total hip replacement are the most difficult. However, the road to full recovery is a long one and PearlDiver data shows evidence there is a period in which the odds of a post-operative fracture decrease dramatically. As shown in Table 2, within 30 days of the total hip arthroplasty, 85% of fractures taking place in the first six months have already occurred. So for patients working to move without walkers or crutches, after day 30 there is only a 15% chance of fracture of the femur within the next five months.
Table 2: Distribution of Time to Diagnosis
|
2 |
5 |
30 |
90 |
120 |
180 |
% of Total |
25.9% |
25.8% |
33.7% |
10.5% |
1.7% |
2.5% |
U.S. |
1,111 |
1,105 |
1,444 |
450 |
72 |
106 |
Source: PearlDiver Patient Records Database, 2004-2007
These numbers show a strong potential for fractures present from the arthroplasty procedure itself, manifesting during the very early stages of the recovery process. This is true especially within the first two to five days. Yet looking at the numbers of patients under the age of 65 is still a very broad group. Using the PearlDiver Database, we can look deeper in to which patients are most likely to encounter such a situation.
Table 3: Time Distribution of Fractures Diagnosed within 180 Days of Procedure
|
Number of Days |
Age |
2 |
5 |
15 |
30 |
60 |
90 |
120 |
180 |
Total |
Under 15 |
0.0% |
0.0% |
0.0% |
0.0% |
0.0% |
100.0% |
0.0% |
0.0% |
100.0% |
15 to 24 |
0.0% |
75.0% |
0.0% |
25.0% |
0.0% |
0.0% |
0.0% |
0.0% |
100.0% |
25 to 34 |
17.6% |
35.3% |
17.6% |
11.8% |
11.8% |
0.0% |
5.9% |
0.0% |
100.0% |
35 to 44 |
14.9% |
32.8% |
19.4% |
11.9% |
6.0% |
4.5% |
7.5% |
3.0% |
100.0% |
45 to 54 |
21.1% |
30.0% |
18.8% |
11.7% |
8.1% |
4.0% |
2.2% |
4.0% |
100.0% |
55 to 64 |
30.3% |
22.6% |
23.6% |
10.8% |
6.3% |
3.1% |
1.0% |
2.4% |
100.0% |
Source: PearlDiver Patient Records Database, 2004-2007
Table 3 illustrates the time during which these fractures are occurring throughout the first six months of the patient’s recovery process. These numbers represent the percentage of fractures from the total occurring within the given time period. As shown, patients most likely to be diagnosed with a femoral fracture after a THR within the first two days are those in the 55 to 64 age group. Yet, three-quarters of all fractures that will occur within the first six months for those ages 15 to 24 happen on days three through five.
But that time period of day three through day five tells a very interesting story. First, aside from the under 15 category, the 55 to 64 age group posts the least number of fractures on a percentage basis. Taking a closer look at patients between the ages of 25 to 64, the results get increasingly interesting as we analyze the post-procedure timeline. One of the most interesting details is the total of the first five days post-procedure. On a percentage basis, our expectation was that the age group of 55 to 64 would emerge as the clear leader, with the most fractures. Yet this is, in fact, not the case. For this age group, the percentage of fractures diagnosed within the first five days was 52.9%, an astounding number, but just as astounding was that patients age 15 to 24 suffered the same percentage within this time frame.
The age group with the lowest percentage of fractures within the first five days was actually those between the ages of 35 to 44, which still boasted a five-day percentage of 47.8% of the total fractures diagnosed. Within a 15-day time frame, this group still saw the least number of fractures on a percentage basis, yet 67.2% of total number of fractures had already taken place.
From a gender standpoint, the numbers also show significant differences. Within the PearlDiver patient sample, 52% of THR patients were male and 48% female. Yet the number of fractures reported at 180 days was 44% male and 56% female.
Looking more closely as we did previously with age, at five days the results were even more extreme. The number of males diagnosed with a fracture in the first five days was 42%, while the number of females was 58%. Thus these numbers would indicate that female patients are more likely to have a femoral fracture during or shortly after a THR.
From this information, it would be expected that patients undergoing a total hip revision would follow the same demographics and even be at greater risk for a fracture during early recovery. However, the results actually contrast with these expectations.
Table 4: Percentage of Revision Patients Diagnosed With Fracture
|
2004 |
2005 |
2006 |
2007 |
Total |
Within 2 Days |
1.0% |
0.5% |
1.0% |
0.3% |
0.7% |
Male |
0.7% |
0.5% |
1.5% |
0.6% |
0.8% |
Female |
1.2% |
0.5% |
0.5% |
0.0% |
0.5% |
Within 5 Days |
1.6% |
1.8% |
1.2% |
1.8% |
1.6% |
Male |
1.4% |
1.0% |
2.0% |
2.9% |
1.8% |
Female |
1.8% |
2.6% |
0.5% |
0.6% |
1.3% |
Within 30 Days |
4.9% |
4.3% |
4.1% |
4.5% |
4.4% |
Male |
4.3% |
3.0% |
5.5% |
3.5% |
4.1% |
Female |
5.4% |
5.6% |
2.8% |
5.5% |
4.7% |
Within 180 Days |
6.8% |
7.1% |
7.5% |
6.8% |
7.1% |
Male |
6.5% |
5.6% |
9.5% |
5.8% |
6.9% |
Female |
7.1% |
8.7% |
5.6% |
7.9% |
7.3% |
Source: PearlDiver Patient Records Database, 2004-2007
For THR patients, the percentage diagnosed with a femoral fracture within the first five days was nearly 2% of patients and, while as of 2007 the number of revision patients spiked to 1.8%, that is still less than the percentage of THR patients and contrary to what would be expected. Also in this five-day time period, as of 2007 the number of male patients diagnosed with a fracture significantly exceeded the number of female patients, which also diverges from expectations.
Unfortunately, one of the greater contrasts between THR patients and patients undergoing a total hip revision is the long-term trend over 180 days which falls more in line with expectations. While patients diagnosed with a fracture after a THR during this time period averaged 4.1%, patients after a revision averaged 7.1%. Also, the gender percentages change dramatically over time for the revision group as well. While after five days the number of females diagnosed had been exceptionally small, by 30-days post-procedure, females show to be 2% more likely to have suffered a fracture.
With this information we can better gage which patients are more likely to be predisposed to fractures after a THR, but these numbers indicate something more as well. From the time period of 2004 through 2007 the overall percentage of fractures diagnosed after a THR within five days post-procedure actually dropped only slightly from 2% to 1.9%, and the percentage of fractures diagnosed within 180 days dropped from 4.2% to 3.5%.
This tells us that progress has been made for patients during the rehabilitation process, but unseen fractures taking place during the procedure, and fractures occurring very shortly after the procedure continue on with little improvement. Looking forward, the more we know about the potential for such fractures to occur, the more warning signs may be there.