Letter to the Editor: How Accurate Are the Surgical Risk Preoperative Assessment System (SURPAS) Universal Calculators in Total Joint Arthroplasty?

To the Editor,

In their article in Clinical Orthopaedics and Related Research®, Trickey et al. [4] assessed the accuracy of the Surgical Risk Preoperative Assessment System model predictions of 30-day postoperative morbidity and mortality for patients undergoing non-emergent THA and TKA. The authors found that venous thromboembolism, including deep vein thrombosis and pulmonary embolism, occurred at higher rates in patients undergoing total joint arthroplasty (TJA) than predicted by the Surgical Risk Preoperative Assessment System model.

We theorize that observed venous thromboembolism rates were higher than predicted because of the inclusion of pulmonary emboli in this composite endpoint. We believe overly sensitive imaging modalities—including multi-detector row CT (MDCT)—are increasingly identifying clinically insignificant emboli that do not necessary warrant therapeutic anticoagulation, such as isolated subsegmental pulmonary emboli [6]. This may explain why the incidence of non-fatal pulmonary embolism after TJA has changed only negligibly over the past two decades, despite widespread use of aggressive chemical prophylaxis [5].

To illustrate how the rates of isolated subsegmental pulmonary emboli detection may impact model predictions, we determined the incidence of isolated subsegmental pulmonary embolism after primary elective TJA in all patients for whom chest CT scans were ordered at an academic tertiary referral center. This inquiry began as part of an institutional quality-improvement project that we subsequently received institutional review board approval to share as research; the data we share here are preliminary, and they have not been published elsewhere. We are choosing to share these in a Letter to the Editor rather than in a peer-reviewed publication because the finalized version of our study will no longer stratify chest CT scan results by the level of the emboli in the pulmonary arteries due to resultingly low sample sizes, which precluded meaningful group comparisons. Between March 2015 and December 2017, we identified 232 patients who underwent primary THA or TKA who also underwent MDCT within 30 days of surgery for a suspected acute pulmonary embolism. Final radiology reports were used to characterize each as positive or negative for pulmonary embolism; positive scan results were further categorized as either subsegmental, higher-level (segmental, lobar, or main), or both. Incidences and 95% confidence intervals were derived in Stata/IC 15.1 (StataCorp LP: College Station, TX, USA) [3].

Among these patients, 78% (182 of 232; 95% CI, 73%-84%) had a negative MDCT result for pulmonary embolism. A total of 7% (15 of 232; 95% CI, 3%-10%) of patients had isolated subsegmental pulmonary emboli, whereas 4% (10 of 232; 95% CI, 2%-7%) had exclusively segmental or higher-level findings and 11% (25 of 232; 95% CI, 7%-15%) had both subsegmental and higher-level findings (Table 1).

Table 1.

Incidence of pulmonary emboli on chest MDCT after primary elective TJA, stratified by the location of the pulmonary embolism in the pulmonary arteries (n = 232 patients)

CT scan result	Percentage (n)	95% CI
Negative for pulmonary embolism	78% (182)	73%-84%
Positive for pulmonary embolism	22% (50)	16%-27%
Isolated subsegmental pulmonary embolism	8% (15)	3%-10%
Isolated segmental or higher-level pulmonary embolism	4% (10)	2%-7%
Both	11% (25)	7%-15%

Thus, about one-third of patients with positive MDCT results for pulmonary embolism had isolated subsegmental pulmonary emboli after TJA at our institution. Given that the clinical relevance of isolated subsegmental pulmonary emboli is dubious, their inclusion in both contemporary predictive models and treatment algorithms may lead to unwarranted anticoagulation therapy postoperatively that may lead to otherwise avoidable iatrogenic complications [2]. For instance, Donato et al. [1] showed that patients who were not treated with therapeutic anticoagulation for isolated subsegmental pulmonary embolism experienced no complications; in contrast, more than 10% of patients with isolated subsegmental pulmonary emboli experienced hemorrhage or recurrent pulmonary embolism after receiving therapeutic anticoagulation.

To improve patient safety and reduce complications after TJA, the treatment of isolated subsegmental pulmonary emboli should be balanced against the inherent risk of anticoagulation in this patient population.