NEW ORLEANS – Comanagement of orthopedic inpatients by an internist or hospitalist can improve outcomes in myriad ways, Mary Anderson Wallace, MD, said at the annual meeting of the American College of Physicians.

She focused on patients undergoing total hip or knee arthroplasty (THA/TKA). In 2014, there were 400,000 of them under Medicare alone, accounting for $7 billion in hospitalization costs and nearly as much again in the cost of related postdischarge care.

So, changes in management that improve key outcomes in this population by even a small increment reap huge benefits when spread across this enormous patient population, noted Dr. Wallace, an internist and hospitalist at the University of Colorado, Denver.

Among the examples she highlighted where comanagement can have a favorable impact were optimization of perioperative pain management pathways; how to handle the use of disease-modifying antirheumatic drugs (DMARDs) in patients undergoing THA/TKA; the latest thinking on the appropriateness of low-dose aspirin for deep vein thrombosis (DVT) prophylaxis; a simple way to predict postop delirium in older individuals without known dementia; how to decide which postoperative fevers warrant a costly infectious disease workup; and the optimal wait time from arrival at the hospital with a fractured hip and THA.

These are all issues where a well-informed internist/hospitalist can be of enormous assistance to a busy orthopedic surgeon in providing high-value care, she explained.

Optimizing perioperative pain management pathways

As of 2015, orthopedists ranked as the third highest prescribers of opioids. Impressively, a retrospective cohort study of 641,941 opioid-naive, privately insured patients undergoing 1 of 11 types of surgery demonstrated that TKA was associated with a 5.1-fold increased risk for subsequent chronic opioid use in the first year after surgery, compared with 18 million opioid-naive nonsurgical controls. Indeed, TKA was the highest-risk of the 11 surgical procedures examined (JAMA Intern Med. 2016 Sep 1;176[9]:1286-93).

Another study that points to a need to develop best practices for opioid prescribing in orthopedic surgery – and other types of surgery as well – was a systematic review of six studies of patients who received prescription opioid analgesics in conjunction with seven types of surgery.

Opioid oversupply was identified as a clear problem: 67%-92% of patients in the six studies reported unused opioids. Up to 71% of opioid tablets went unused, mainly because patients felt they’d achieved adequate pain control and didn’t need them. Rates of safe disposal of unused opioids were in the single digits, suggesting that overprescribing provides a large potential reservoir of opioids that can be diverted to nonmedical use (JAMA Surg. 2017 Nov 1;152[11]:1066-71).

Moreover, a recent retrospective study of more than 1 million opioid-naive patients undergoing surgery showed that 56% of them received postoperative opioids. And each additional week of use was associated with a 44% increase in the relative risk of the composite endpoint of opioid dependence, abuse, or overdose. Duration of opioid use was a stronger predictor of this adverse outcome than was dosage (BMJ. 2018 Jan 17;360:j5790).

Other studies have shown that multimodal analgesia is utilized in only 25%-50% of surgical patients, even though it is considered the standard of care. Only 20% of patients undergoing THA/TKA receive peripheral nerve and neuraxial blocks. So, there is an opportunity for optimization of perioperative pain management pathways in orthopedic surgery patients, including avoidance of unnecessary p.r.n. prescribing, to favorably impact the national opioid epidemic, Dr. Wallace observed.

A surprise benefit of multimodal pain management that includes acetaminophen and a nonsteroidal anti-inflammatory agent is that it markedly reduces the incidence of postoperative fevers after total joint arthroplasty, compared with opioid-based pain management.

That was demonstrated in a retrospective study of 2,417 THA/TKAs in which multimodal pain management was used, and 1,484 procedures that relied on opioid-based pain relief. All of the operations were performed by the same three orthopedic surgeons. Only 5% of patients in the multimodal pain management group developed a fever greater than 38.5 degrees Celsius during the first 5 postoperative days, compared with 25% of those in the opioid-based analgesia group.

Moreover, an infectious disease workup was ordered in 2% of the multimodal analgesia group, versus 10% in the opioid-based pain management cohort, with no difference in the positive workup rates between the two groups (Clin Orthop Relat Res. 2014 May;472[5]:1489-95).

“It’s interesting that multimodal pain management has the side effect of putting you in a better position to practice high-value care, with less fever and fewer infectious disease workups,” Dr. Wallace said.

Perioperative management of DMARDs

Recent joint guidelines from the American College of Rheumatology and American Association of Hip and Knee Surgeons specifically address this issue in patients undergoing elective joint replacement (Arthritis Rheumatol. 2017 Aug;69[8]:1538-51).

“The quality of evidence isn’t high, but at least it’s a starting point,” Dr. Wallace said.

Patients with rheumatoid arthritis, systemic lupus erythematosus, spondyloarthropathies, and other rheumatic diseases who are on methotrexate or other nonbiologic DMARDs should be maintained on their current dose, according to the guidelines.

In contrast, all biologic agents should be withheld prior to surgery, which should be scheduled to coincide with the end of the dosing cycle for that specific biologic. The biologic agent should be resumed only once adequate wound healing has occurred, typically about 14 days post-THA/TKA.

Patients on daily glucocorticoids should continue on their current dose; supraphysiologic stress dosing is to be avoided.

Low-dose aspirin for VTE prophylaxis

“It seems like nothing has been such an enduring controversy in the comanagement literature as the question of whether aspirin is an effective prophylactic agent for prevention of DVT post THA/TKA,” according to Dr. Wallace.

She noted that in the space of just 4 years between the eighth and ninth editions of the American College of Chest Physician guidelines, that organization underwent a 180-degree reversal on the issue – whipsawing from a grade 1A recommendation against aspirin in 2008 to a 1B recommendation for it in 2012.

The literature is increasingly supportive of the use of aspirin for venous thromboembolism (VTE) prophylaxis in low-risk THA/TKA patients. Separate guidelines from the American Academy of Orthopaedic Surgeons (AAOS) and the Surgical Care Improvement Project, as well as the chest physicians, support the practice.

The hitch is that there is as yet no single validated risk-stratification protocol. AAOS recommends 325 mg of aspirin twice daily for 6 weeks. But a prospective crossover study of more than 4,600 total joint arthroplasty patients conducted by investigators at Thomas Jefferson University in Philadelphia showed that 81 mg BID for 4 weeks was just as effective as was 325 mg b.i.d., albeit with an incidence of GI bleeding that to their surprise wasn’t significantly lower (J Bone Joint Surg Am. 2017 Jan 18;99[2]:91-8).

Dr. Wallace anticipates definitive answers on VTE prophylaxis to come from the ongoing Patient-Centered Outcomes Research Institute-supported PEPPER (Comparative Effectiveness of Pulmonary Embolism Prevention After Hip and Knee Replacement) trial. In that study, roughly 25,000 patients undergoing THA/TKA are being randomized to prophylactic aspirin at 81 mg b.i.d., warfarin at an International Normalized Ratio of 2.0, or rivaroxaban (Xarelto). Endpoints include mortality, VTE, and bleeding. Results are expected in 2021.

Preoperative prediction of postop delirium

Unrecognized preoperative cognitive impairment in older patients without dementia who are undergoing THA/TKA is a common and powerful risk factor for postop delirium and other complications, as demonstrated recently by investigators at Harvard University and affiliated hospitals.

They had 211 patients aged 65 or older, none with known dementia, take the Mini-Cog screening test prior to their surgery. Fully 24% had probable cognitive impairment as reflected in a score of 2 or less out of a possible 5 points on this simple test, which consists of a three-word recall and clock drawing.

“I was very surprised at this high rate. These are patients who are at risk for delirium in the hospital when you’re taking care of them,” Dr. Wallace observed.

In the Harvard study, the incidence of postop delirium was 21% in patients with a Mini-Cog score of 2 or less, compared with 7% who scored 3-5, for an odds ratio of 4.5 in an age-adjusted multivariate analysis. Moreover, 67% of the low scorers were discharged somewhere other than home, in contrast to 34% of patients with a preop Mini-Cog score of 3-5, for an adjusted 3.9-fold increased risk. The group with a Mini-Cog score of 2 or less also had a significantly longer hospital length of stay (Anesthesiology. 2017 Nov;127[5]:765-74).

Perioperative gabapentin is often added to the medication regimen of older surgical patients to reduce postop delirium. The latest evidence indicates that doesn’t work, as demonstrated in a recent 697-patient randomized trial. The incidence of delirium during the first 3 days post surgery as measured by the Confusion Assessment Method was 22.4% in patients randomized to 900 mg of gabapentin per day, and 20.8% with placebo. Nearly 200 participants had THA or TKA, and in that subgroup, there was an even stronger – albeit still not statistically significant – trend for a higher delirium rate with gabapentin than with placebo (Anesthesiology. 2017 Oct.127[4]:633-44).

“Think twice about adding gabapentin to the pain regimen for THA/TKA/spine patients for the purpose of preventing postop delirium,” she advised.

When is postop fever a concern?

Up to half of patients develop fever early after THA/TKA. In most cases, this is a self-limited ancillary effect of cytokine release, with the temperature peaking on postop day 1-2.

Three strong predictors of a positive infectious disease workup are fever after postop day 3, with an associated 23.3-fold increased risk; multiple days of fever, with an odds ratio of 8.6; and a maximum temperature greater than 39.0 degrees Celsius, with a 2.4-fold increased risk. In this 7-year-old study, the cost of infectious disease workup per change in patient management was a hefty $8,209 (J Arthroplasty. 2010 Sep;25[6 Suppl]:43-8).

A retrospective study of nearly 125,000 THA/TKA patients in the American College of Surgeons National Surgical Quality Improvement Program database has important implications for clinical surveillance for postop adverse events. Stroke occurred early, on median postop day 1. The median time of acute MI and pulmonary embolism was postop day 3, and pneumonia day 4.

The key take-home message was that the median time to DVT was postop day 6, by which point most patients had been discharged. Thus, 60% of postoperative DVTs occurred after discharge. And the time to diagnosis of DVT differed markedly by surgical procedure: The median day of diagnosis was day 5 in TKA patients, compared with day 13 for THA patients. Sixty-eight percent of urinary tract infections occurred post discharge. Sepsis occurred on median day 10 post surgery, surgical site infections on day 17 (Clin Orthop Relat Res. 2017 Dec;475[12]:2952-9).

In light of ever-shortening hospital lengths of stay, Dr. Wallace noted, the findings underscore the importance of comprehensive predischarge patient counseling.

Optimal time window for hip fracture surgery

AAOS guidelines recommend that hip fracture surgery should take place within 48 hours, assuming medical comorbidities are stabilized, because complication rates go up with longer wait times.

But that is controversial. A University of Toronto retrospective cohort study of 42,430 adults with hip fracture treated at 72 Canadian hospitals during 2009-2014 found that the inflection point was 24 hours. Among 13,731 patients whose elapsed time from hospital arrival to surgery was 24 hours or less, 30-day mortality was 5.8%, significantly less than the 6.5% rate in an equal number of propensity score–matched patients with a longer wait time.

The 90- and 365-day mortality rates in the patients who received surgery within 24 hours were 10.7% and 19.3%, both significantly lower than the 12.0% and 21.6% figures in patients with longer wait times.

For the 30-day composite outcome of death, myocardial infarction, pulmonary embolism, DVT, or pneumonia, the rates were 10.1% and 12.2% – again, statistically significant and clinically meaningful. The 90- and 365-day composite outcomes followed suit (JAMA. 2017 Nov 28;318[20]:1994-2003).

But the Canadian study won’t be the final word. The ongoing international multicenter HIP ATTACK (Hip Fracture Accelerated Surgical Treatment and Care Track) trial is comparing outcomes in 3,000 patients randomized to hip fracture surgery within 6 hours versus 24 hours. Endpoints include mortality, myocardial infarction, pulmonary embolism, pneumonia, stroke, sepsis, and life-threatening and major bleeding.

Dr. Wallace reported having no financial conflicts regarding her presentation.

bjancin@mdedge.com

Internist comanagement of orthopedic inpatients boosts outcomes