1 Introduction

Over 1 million total joint replacements take place annually in the United States.¹ While total hip arthroplasty (THA) is considered an effective surgery, complications can have severe repercussions due to older patients limited physiologic reserve.² Unforeseen medical problems can lead to prolonged convalescence, poor post-operative outcomes, and increased mortality.^3,4

The medical world has also been shaken by the significant opioid crisis and increasing need for pain control therapies, medications, or interventions. The U.S. outpaces all other nations in opioid prescribing, with the third-highest prescribers of narcotics being orthopedic surgeons.⁵ Recent research has focused on the risk implications due to medical conditions like diabetes, obesity, or cardiac disease. New national initiatives prioritize the use of alternative pain measures, partially through optimized indications and reduced overall complications.⁶ A focus on local anesthesia, regional blocks, and the use of multi-modal pain control is advocated by all specialty and regulatory organizations. Nevertheless, the demand for opioids continues to rise, as well as the associated complications of opioid overuse and abuse.^7,8 To appropriately serve our patients, it becomes imperative to understand indications and factors or comorbidities that drive outcomes and complications.

When considering opioid overuse, COPD is not the typical patient population to highlight. There are approximately 16 million people living with chronic obstructive pulmonary disease (COPD) in the United States.³ Although COPD primarily affects the lungs, comorbidities impacting other systems including psychiatric/neurological, cardiovascular, musculoskeletal diseases and disorders, and cancer⁴ are extremely common, as well as overall increased morbidity and mortality.⁹ This extensive multi-system involvement costs the average patient over $22,000 additional per episode, resulting in a substantial economic burden on the national healthcare system.¹⁰

Much research has gone into the epidemiology and mechanisms of COPD and the associated surgical implications. While previous studies document significantly increased risk for complications such as increased length of stay, infection, or readmission in THA,^11,12 there is a scarcity of literature investigating the association of COPD and peri-operative risks in THA or postoperative opioid use in these patients. Considering both the economic and health-associated impact of COPD, the current study attempts to determine the specific risks associated with elective THA, especially in relation to opioid use as well as associated complications.

2 Methods

Study Design: A retrospective follow-up analysis of Medicare patients using the PearlDiver Patient Records Database (http://www.pearldiverinc.com) was used. The national database contains deidentified medical records from Humana Inc. and Medicare, encompassing approximately 25 million records spanning the period between 2007 and 2017.

Patient Cohort: THA patients between 2007 and the first quarter of 2017 were selected using the first instance Current Procedure Terminology (CPT) code CPT-27130, which refers to “Repair, Revision, and/or Reconstruction Procedures on the Pelvis and Hip Joint” (see Appendix A). Then, two cohorts were established: 1) THA patients with COPD and 2) THA patients without COPD. The COPD cohort was determined based on a predefined list of ICD-9-CM codes related to COPD in the COPD Comorbidity Select Tool. COPD is an umbrella term encompassing asthma, bronchitis, and emphysema. Patients with any of these codes in their medical records prior to surgery were included in the COPD cohort. Patients were excluded if they had traumatic hip fractures, multiple myeloma, Paget's disease of the bone, metastatic cancer, achalasia, IBD, cachexia, connective tissue disease/rheumatologic disease, epilepsy, poliomyelitis, myasthenia gravis, Parkinson's, or previous bariatric surgery as these may influence outcomes. Age distribution and gender along with relevant comorbidity data such as tobacco use, obesity, hypertension, peripheral vascular disease, asthma, chronic kidney disease, depression, congestive heart failure, dysglycemia (diabetes, pre-diabetes, and other blood sugar disorders), urinary tract infection, and pneumonia were recorded for each cohort.^4,6 Additionally, the mean Charlson Comorbidity Index (CCI) was computed for each cohort. The CCI was reported in the database for each patient and the mean was calculated from these reported values.

Outcome Measures: Opioid use and medical and implant-related postoperative complications within 30 days of THA were identified with ICD-9-CM and ICD-10-CM codes (refer to Appendix B). Opioid utilization preoperatively was determined through examination of filled opioid prescriptions records within three months of THA.¹³ Postoperatively, opioid utilization was assessed at one, three, six, and twelve-month intervals, and usage for more than three months was considered prolonged.^14–16 Medical complications included surgical site infection, neurological issues, sepsis, deep vein thrombosis (DVT), pneumonia, myocardial issues, complications of the respiratory system, cardiac issues, cardiac arrest, coagulation disorders, mortality, a total of all complications, as well as 30-day readmissions. The study also assessed the rate of COPD exacerbation, defined as a worsening of symptoms or, for this study, patients without COPD who were seen for respiratory complications including shortness of breath, respiratory distress, or pneumonia. Dislocation, prosthetic infection, mechanical failure, periprosthetic fracture, and their aggregate total were also considered as implant related complications. The study further specified revision THA using CPT code (see Appendix A) and calculated the median time from primary to revision THA.

Statistical Analysis: Using Chi-squared analysis the descriptive statistics and comorbidity prevalence of each cohort was compared. The postoperative outcome occurrence and opioid use odds ratios were calculated counting the number of events within 30 days of THA. Adjusted odds ratios were calculated using multivariable logistic regression, while controlling for age, sex, medical comorbidity, and preoperative opioid usage. Using Cox-proportional hazards modeling and Kaplan-Meier survival analysis with a log-rank test the risk of THA revision was compared. All analyses were performed using the PearlDiver software, Version 1.1.442 (RStudio Inc., Boston MA) with a significance level, α value, of 0.05.

3 Results

From 2007 to the first quarter of 2017, 22,931 THA COPD patients (inclusive of diagnoses such as COPD, emphysema, bronchitis, and asthma) and 60,513 THA non-COPD patients were studied (Table 1). The age and sex distribution can be found in Table 1. The prevalence of all the observed comorbidities was higher in COPD patients (p = 0.0223 for chronic kidney disease, p< 0.0001 for others) with a mean CCI of 3.34 compared to 1.40 in non-COPD patients. Common comorbidities for COPD patients included tobacco use (51.8%), hypertension (92.4%), urinary tract infection (53.5%), and dysglycemia (57.6%).

3.1

3.1 THA patients with COPD: opioid use and medical complications

Multivariate regression adjusted for age, gender, race, and CCI found that THA patients with COPD had significantly greater opioid utilization in comparison to those undergoing THA without COPD (Fig. 1). After adjusting for preoperative opioid use, multivariate analysis showed THA COPD patients were more likely to pick up opioid prescriptions at 1, 3, 6, and 12 months (aOR = 1.18, 95% CI 1.13–1.23, p< 0.0001, aOR = 1.17, 95% CI 1.11–1.23, p< 0.0001, aOR = 1.22, 95% CI 1.16–1.29, p< 0.0001, aOR = 1.30, 95% CI 1.22–1.38, p< 0.0001 respectively) (Table 2). At the same time, THA COPD patients had an increased rate of readmission within 30-days (10.6% vs. 7.1%, 95% CI 1.19–1.34, aOR = 1.26, p< 0.0001) and increased 30-day total medical complications (9.9% vs. 6.3%, 95% CI 1.20–1.36, aOR = 1.28, p< 0.0001), when compared to THA non-COPD patients (Table 2). Risk of surgical site infection (2.5% vs. 1.7%, 95% CI 1.07–1.35, aOR = 1.20, p = 0.0025), respiratory-related (2.7% vs. 0.9%, 95% CI 1.87–2.43, aOR = 2.13, p< 0.0001), and pneumonia (0.6% vs. 0.2%, 95% CI 1.86–2.60, aOR = 2.20, p< 0.0001) were higher in COPD patients undergoing THA. The rate of COPD exacerbation after THA in COPD patients was 8.4% within 30 days of the surgery. Mortality risk within 30 days of THA was 0.2% and not different between THA patients with or without COPD (aOR = 1.13, 95% CI 0.96–1.33, p = 0.5101).

Close

Fig. 1 Post-operative opioid use in THA patients with and without COPD at 1, 3, 6, and 12-months following THA.

Export to PPT

Table 2 Medically related complications for total hip replacement patients with COPD.

	THA w/o COPD		THA w COPD		Adjusted Odds Ratio	95%CI	p-value
	n	%	n	%
30 day Readmission	4304	7.1	2438	10.6	1.26	1.19–1.34	< 0.0001
30 day COPD Exacerbations/Respiratory Complications a	1917	3.2	471	2.1	6.24	5.58–6.98	< 0.0001
30 day Total Medical Complications	3791	6.3	2270	9.9	1.28	1.20–1.36	< 0.0001
Surgical Site Infection	1034	1.7	570	2.5	1.20	1.07–1.35	0.0025
DVT	1712	2.8	828	3.6	1.05	0.96–1.16	0.2830
Respiratory	553	0.9	629	2.7	2.13	1.87–2.43	< 0.0001
Sepsis	310	0.5	206	0.9	1.16	0.95–1.42	0.1453
Myocardial	370	0.6	206	0.9	1.18	0.97–1.43	0.0895
Pneumonia	118	0.2	144	0.6	2.20	1.86–2.60	< 0.0001
Cardiac Arrest	105	0.2	56	0.2	1.10	0.77–1.58	0.5968
Cardiac	99	0.2	40	0.2	1.01	0.66–1.19	0.9630
Death	100	0.2	55	0.2	1.13	0.96–1.33	0.5101
Opioid Utilization
1 Month Preoperatively	11,649	19.3	6418	28.0	–	–	–
1 Month Postoperatively	11,464	19.0	6279	27.4	1.18	1.13–1.23	< 0.0001
3 Months Postoperatively	7046	11.7	4330	18.9	1.17	1.11–1.23	< 0.0001
6 Months Postoperatively	5633	9.3	3663	16.0	1.22	1.16–1.29	< 0.0001
12 Months Postoperatively	4303	7.1	2910	12.7	1.30	1.22–1.38	< 0.0001

a In patients without COPD, respiratory complications were considered to be similar to COPD exacerbations. These respiratory complications included shortness of breath, respiratory distress, or pneumonia.

4 Discussion

COPD is a modifiable condition that can be optimized with medications, lifestyle management, and careful monitoring from a pulmonologist. We have previously investigated the correlation of COPD patients with the risk of complications after total knee arthroplasty (TKA).¹⁷ Patients with COPD were at a higher risk for postoperative medical complications and increased opiate usage at 3, 6, and 12 months after TKA. We have also looked at COPD in total hip arthroplasty but had no information on opioid use and the associated comorbidities.¹⁸ COPD is not typically one of the at-risk conditions considered in opioid overuse or abuse.

Unfortunately, COPD patients often carry many other risk factors that are assumed to take precedence over the COPD diagnosis. The relationship between COPD and THA is poorly explored, especially in the rapidly expanding Medicare population. To address the shortcomings of available research, current work hypothesized that COPD not only increased the overall risk of THA but also increased the need for opioid consumption and possibly overuse and abuse. Data collection focused on pre-surgical and post-surgical opioid use and medical complications such as surgical site infection, DVT, sepsis, cardiac arrest, and death.

Currently, few large-scale analyses of COPD in total joints are present in the literature.^19,20 Previous studies were conducted in Taiwan and were limited due to the utilization of a single-payer healthcare system and a small patient population.^21,22 We have previously published on a different patient cohort focusing on private insurers and both total knee and total hip outcomes in COPD patients, showing overall increased risks. The present study is the largest of its kind and builds upon results from our previous studies, effectively showing that patients diagnosed with COPD have worse outcomes after THA compared to patients without COPD. More importantly, the presented analysis is one of few evaluating opioid use in COPD patients after total hip arthroplasty. Such information may allow for improved planning and education, focus on alternative treatments and stringent indications around post-operative pain management.

After previously determining that COPD patients following TKA had increased opioid usage at 3, 6, and 12 months, we expanded our time frame to include opioid usage at 1 month in the current study. Interestingly, patients in the COPD cohort after THA were significantly more likely to be taking opioid medication at 1, 3, 6, and 12 months after surgery (p< 0.0001 for all timepoints) even after controlling for preoperative opioid usage. Because osteoarthritis is a chronic and painful condition, patients who receive total joint arthroplasty (TJA) are often on opioid medications for more extended periods of time than other populations.²³ Unfortunately, COPD patients have higher exposure to opioid medications at baseline. In a 2016 study by Vozoris et al., new opioid prescriptions were given to 68.2% of patients with COPD, leading to significantly higher rates of hospitalizations, morbidity, and mortality.²⁴ Such prevalence could potentially lead to increased opioid addiction and withdrawal symptoms, making TJA recovery more complicated, and result in the increased usage seen in this study. As such, aggressive screening for opioid use in the COPD population before TJA should be considered, including devising a plan for a multimodal postoperative pain control regimen and close follow-up with a primary care physician.

COPD patients undergoing THA had increased readmission rates, postoperative medical complications, implant-related infections, and increased rates of revision THA when compared to their healthy counterparts. Such results were expected and fell in line with our previous research. Patients with COPD have increased hospital visits due to exacerbations, which may contribute to the medical complications. They are also frequently treated with steroids for these exacerbations. Steroid treatment results in immunosuppression and increased risk of infection and can also have negative affect on bone quality.^25,26 This may explain then why COPD patients were more likely to have infections and need revisions, they potentially have increased risk of infection and worse bone quality for implants. Considering the impact and cost of poor outcomes in arthroplasty, understanding COPD's role in each of these complications is paramount to appropriate preoperative patient selection and medical optimization.

As one might expect, the patients with COPD had an increased likelihood to contract pneumonia and other respiratory problems following THA. However, despite a trend toward myocardial or cardiac complications, as would be expected, no significant relationship was observed in the present study. COPD affects the pulmonary and cardiovascular systems due to obstructed air exchange in the lungs which leads to a reduction in oxygen and subsequent downstream effects across all systems.²⁷ A significantly higher rate of surgical site infections was reported in patients with COPD. The COPD cohort was more likely to have implant complications, such as periprosthetic fracture, prosthetic infection, mechanical loosening, and dislocation. As a result, COPD was correlated with a higher incidence of revision total hip arthroplasty. In 2018, Klasan et al. found that COPD patients who received lower limb arthroplasty had higher CRP levels and leukocyte counts, which could possibly reflect a low-level inflammatory reaction or be a result of the presence of a chronic disease.¹⁹ The higher CRP and leukocyte counts seen in COPD patients may dysregulate the immune system and increase stress on the body, which could lead to poor soft tissue healing and decreased implant osseointegration.²⁸ To minimize the impact of COPD, physicians should make sure that these patients are being closely followed by their primary care or a pulmonologist to ensure medical optimization. Patients may also benefit from preoperative education to discuss lifestyle changes that may decrease their risk of complications, including alternative pain management and opioid use reduction, anti-inflammatory diets, smoking cessation, weight loss, and stress management.

Understanding the impact COPD has on the outcomes following THA can have a significant impact on clinical decision making in these patients. Patient with COPD have been shown to have some of the lowest function and quality of life scores before THA when compared to baseline or patients with other comorbidities. They do however have gain in function comparable to other comorbid conditions up to 1 year post THA.²⁹ While these patients have increased complications and increased opioid usage, there is still utility in improving patient function. This then stresses how to treat these patients. Patients with COPD may be better suited for inpatient procedures. Many institutions have different criteria to qualify for outpatient THA. At our institution comorbid conditions such as COPD is an exclusion criterion. With the reported increased in complication and opioid use this would be recommended for other institutions as well.

While this study provided valuable insight into how COPD can affect the outcomes of THA, it does have significant limitations. Unfortunately, “big data” studies are plagued by the problems inherent to database research. These include technical issues, collection bias, the observational nature of the method, and follow-up bias. Yet, many recent high-impact studies have used PearlDiver to produce significant, reliable, and well-accepted results in the field of adult reconstruction surgery.^30,31 To optimize the validity of current results, the study parameters and methodology were scrutinized by a university statistician and followed the previous methodology of reference publications. Still, while PearlDiver uses sophisticated algorithms for selection, the database cannot exclude patients who underwent THA for hip fractures. Additionally, COPD patients are more likely to have osteoporosis and have resulting fractures when compared to their non-COPD counterparts.³² This evidence could alter the interpretation of our results, especially in the COPD cohort. The two study groups have differences as well. The COPD patients in this study are older, have a higher percentage of females, and have increased rates of many comorbidities, which could possibly confound our results. However, we controlled for age, sex, and comorbidity status in our statistical analysis.

There are some limitations to the study cohort itself. COPD-related diagnoses may be grouped together, which can possibly lead to a higher rate of COPD prevalence in the current cohort compared to previously reported literature. However, noting recent studies that have found a 13.6% prevalence of COPD and 18.9% prevalence of asthma in THA patients, our findings support similar trends.²² Additionally, while the rates of COPD, dysglycemia, and 30-day readmissions are higher than expected, multiple studies have shown similar values to those presented in this manuscript.³³ Type II diabetes is a common form of dysglycemia – nevertheless, it is also important to note that dysglycemia, as encoded in the PearlDiver patient selection tools, also includes patients with pre-diabetes, clinically low blood sugar, and other blood glucose derangements. This could lead to a higher prevalence of blood glucose abnormalities compared to other studies that focus on diabetes specifically. Recent estimates from Cappozi et al. have shown that over 50% of adults have some sort of blood sugar derangement, supporting current observations.³⁴ Nevertheless, while PearlDiver contains millions of patient records, it only makes up a sample of the United States population. Many private insurance claims are reported by United Healthcare Record.³⁵ This could contribute to variations in the comorbidities and outcomes in this study compared to the general population or previous reports.

5 Conclusions

Higher comorbidity rates along with an increased rate of post-operative medical complications, implant complications, revision THA, and higher use of opioids early and late after THA is associated with COPD in comparison to patients who do not have COPD. Considering the current nationwide opioid crisis, physicians need also to be highly aware of the correlation between COPD and opioid use. Taking steps to optimize all comorbidities and, even more importantly, COPD before surgery, may avert major complications, reduce opioid usage, and improve surgical outcomes.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Author contributions

Jacob Laperche – Conceptualization, data curation, writing, revising, project organization and submission; Caitlin Barrett - Conceptualization, data curation, writing, revising; Jillian Glasser - Conceptualization, data curation, writing, revising; Daniel Yang - Conceptualization, data curation, revising; Nicholas Lemme - Conceptualization, data curation, revising; Dioscaris Garcia - Conceptualization, revising, supervision, project administration; Alan Daniels - Conceptualization, revising, supervision, project administration; Valentin Antoci Jr - Conceptualization, revising, supervision, project administration.