The Journal of Arthroplasty 30 (2015) 1137–1141
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Outcomes Following Primary Total Hip or Knee Arthroplasty in Substance Misusers Matthew J. Best, BS a,1, Leonard T. Buller, MD a,2, Alison K. Klika, MS b, Wael K. Barsoum, MD b,3 a b
University of Miami Miller School of Medicine, Department of Orthopaedic Surgery and Rehabilitation, Miami, Florida Cleveland Clinic Department of Orthopaedic Surgery, Mail Code A41,Cleveland, Ohio
a r t i c l e
i n f o
Article history: Received 20 November 2014 Accepted 31 January 2015 Keywords: total knee arthroplasty total hip arthroplasty drug misuse drug dependence drug abuse
a b s t r a c t The influence of drug misuse on outcomes following primary total hip (THA) or knee (TKA) arthroplasty is poorly understood. The National Hospital Discharge Survey was used to identify patients who underwent primary THA or TKA between 1990 and 2007. Patients were divided into two groups: 1) those with a diagnosis of drug misuse (cannabis, opioids, cocaine, amphetamines, sedatives, inhalants or mixed combinations) (n = 13,163) and 2) those with no diagnosis of misuse (n = 8,366,327). Patients with a diagnosis of drug misuse had longer hospital stays (P b 0.001), nearly eight times the odds of leaving against medical advice (P b 0.001) and five times the mortality rate (P b 0.001). Drug misuse was associated with higher odds (P b 0.001) of complications including postoperative infection, anemia, convulsions, osteomyelitis, and blood transfusion. © 2015 Elsevier Inc. All rights reserved.
Drug misuse, including substance abuse and dependence [1], is a maladaptive pattern of substance use with considerable societal, economic and personal cost [2–5]. The lifetime prevalence of drug abuse and dependence among U.S. adults is estimated at 7.7% and 2.6%, respectively [6]. The risk of secondary osteoarthritis among drug users is thought to be higher than the general population due to increased rates of osteonecrosis of the femoral head and inflammatory arthropathies due to repetitive bacteremia [7,8]. Additionally, when joint arthroplasty is considered, the risk of a septic complication is of concern as injection or inhalation of drugs is felt to increase the risk of infection [9,10]. Despite the rates of drug abuse and dependence in the U.S., few studies have evaluated their influence on outcomes following primary total hip (THA) or knee (TKA) arthroplasty [7,11,12]. Previous groups have found decreased implant survival rates [7], increased rates of infection [12], and high rates of postoperative substance withdrawal delirium [11] in substance abusers who underwent total joint arthroplasty, but these studies were limited by small sample size. The purpose of the present study was to measure the influence of drug misuse on inpatient perioperative outcomes following primary
One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2015.01.052. ⁎ Reprint requests: Leonard T. Buller, MD, Resident physician, University of Miami Department of Orthopaedic Surgery and Rehabilitation, Miami, FL. 1 Medical student, University of Miami Miller School of Medicine, Miami, FL, USA. 2 University of Miami Department of Orthopaedic Surgery and Rehabilitation, Miami, FL, USA. 3 Vice chair, Department of Orthopaedic Surgery, Cleveland Clinic, Cleveland, OH, USA. http://dx.doi.org/10.1016/j.arth.2015.01.052 0883-5403/© 2015 Elsevier Inc. All rights reserved.
THA or TKA using a large national database. The specific aims of the study were to evaluate the relationship between substance abuse and length of hospital stay, discharge disposition, mortality and perioperative complications in patients undergoing primary THA or TKA. The identification of modifiable risk factors associated with complications following total joint arthroplasty may allow surgeons to intervene preoperatively, potentially decreasing complications and improving outcomes. Materials and Methods Data Source The National Hospital Discharge Survey (NHDS) [13], developed by the National Center for Healthcare Statistics Division of the Centers for Disease Control and Prevention (CDC), was used in this study. The NHDS is considered the most comprehensive of all inpatient surgical databases in use today and is the principal database used by the U.S. government for monitoring hospital use [14]. Publicly available, the NHDS provides demographic and medical data for inpatients discharged from non-federal, short stay hospitals [14]. The survey uses International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9CM) [15] codes to classify up to seven discharge diagnoses and up to four procedures that are present at the time of discharge. In addition to medical information, the NHDS collects demographic information (age, gender), expected source of payment (insurance status), length of hospital stay, hospital size, U.S. region, and inpatient outcomes including discharge destination [16]. The NHDS ensures an unbiased national sampling by using a complex three-stage probability design
1138
M.J. Best et al. / The Journal of Arthroplasty 30 (2015) 1137–1141
including: inflation by reciprocals of the probabilities of sample selection, adjustment for no response and population weighting ratio adjustments [14]. This study did not require approval by the institutional review board because the NHDS is a publically available database with no patient identifying information.
Patient Selection All patients admitted to hospitals in the U.S. who underwent primary THA or TKA between 1990 and 2007 were identified using ICD-9-CM codes. Using previously described techniques, discharges with a procedure code (ICD-9-CM) of primary THA (81.51) or TKA (81.54) were identified [17]. Due to National Center for Health Statistics budgetary limitations starting in 2008, the number of hospital surveys was halved, decreasing the precision of the survey data and nearly doubling the relative standard error [18]. Consequently, we chose 2007 as the endpoint of our study. Patients were divided into two groups: 1) those with a diagnosis of drug misuse (ICD-9-CM: 304.00–304.93, 305.20–305.93) and 2) those who did not have a diagnosis of drug misuse. The drug misuse group includes subjects with diagnoses for either substance abuse or substance dependence. Substance abuse is defined as a maladaptive pattern of substance use leading to clinically significant impairment or distress as manifested by one or more of the following, occurring within a 12-month period: 1) recurrent substance use resulting in a failure to fulfill major role obligations at work, school, or home (e.g. repeated absences or poor work performance related to substance use; substancerelated absences, suspensions, or expulsions from school; neglect of children or household), 2) recurrent substance use in situations in which it is physically hazardous (e.g. driving an automobile or operating a machine when impaired) 3) recurrent substance-related legal problems (e.g. arrests for substance-related disorderly conduct) 4) continued substance use despite persistent or recurrent social or interpersonal problems caused or exacerbated by the effects of the substance (e.g. arguments with spouse about consequences of intoxication, physical fights) 5) the symptoms for substance abuse have never met the criteria for substance dependence [1]. In contrast, substance dependence is defined as a maladaptive pattern of substance use leading to clinically significant impairment or distress as manifested by three or more of the following, occurring at any time in the same 12-month period: 1) tolerance, as defined by either of the following: a) a need for markedly increased amounts of the substance to achieve intoxication or desired effect b) markedly diminished effect with continued use of the same amount of the substance, 2) withdrawal, as manifested by either of the following: a) the characteristic withdrawal syndrome for the substance, b) taking the same (or a closely related) substance to relieve or avoid withdrawal symptoms, 3) taking the substance often in larger amounts or over a longer period than was intended, 4) having a persistent desire or unsuccessful efforts to cut down or control substance use [1]. Demographic variables were collected including age, sex and prevalence of comorbidities. The length of hospital stay and discharge destination were determined. The incidence of complications was determined using the complication screening package [19]. The variable “perioperative complication” was created based upon the following (ICD-9-CM): acute postoperative bleeding (998.1), acute postoperative infection (998.5), other operative complication (998.89), acute postoperative anemia (285.1), thrombocytopenia (287.4, 287.5), peripheral vascular complication (997.2), urinary tract infection (599.0), other urinary complication (997.5), acute renal failure (584), acute myocardial infarction (410), pulmonary embolism (415.1), pulmonary insufficiency (518.5), acute deep venous thrombosis (453.4), osteomyelitis (730.0–730.2), cellulitis/abscess formation (682), convulsion (780.39), transfusion of blood (99.0), mechanical complication of internal orthopedic device (996.4, 996.79), infection of internal joint prosthesis (996.77), and other complication of internal joint prosthesis (996.77).
Statistical Analysis Differences between continuous variables were compared using the independent-samples t-test, while the Pearson chi square test was used to compare differences between categorical variables. To determine whether drug abuse was an independent predictor of a negative inhospital outcome, variables present in at least 2% of the population [20] were included in a multivariable binary logistic regression model. The dichotomous variables were 1) presence of any complication, 2) prolonged hospital stay (greater than the 75th percentile of the mean), 3) leaving against medical advice, and 4) in-hospital mortality. Potential confounders were controlled for using a multivariable regression model, to isolate the effect of drug abuse on inpatient outcomes. Covariates accounted for in the regression model included: gender, age, length of stay, hospital bed size, primary source of payment, presence of a complication, and preexisting comorbidities (i.e., diabetes mellitus, hypertension, congestive heart failure, coronary artery disease, atrial fibrillation, previous myocardial infarction, osteoporosis, and rheumatoid arthritis). Odds ratios and confidence intervals were calculated to assess the association between drug abuse and inpatient perioperative complications. Correcting for multiple comparisons, a Pvalue b0.001 was used to define statistical significance, as previously described [21]. All data were analyzed using the software-statistical package for social sciences [SPSS] version 20 (Chicago, IL, USA). Source of Funding No external funding source was used for the conduct of this study. Results A cohort representative of 8,379,490 patients who underwent primary THA or TKA between 1990 and 2007 was identified (Table 1). Of the total cohort, 13,163 patients had a diagnosis of drug misuse, while 8,366,327 patients had no diagnosis of drug misuse. The drug misuse group was younger (51.6 ± 12.9 years compared to 67.3 ± 11.7 years; P b 0.001), had longer hospital stays (5.3 ± 4.0 days compared to 5.1 ± 4.2 days; P b 0.001), and had a higher rate of patients who left against medical advice (0.4% compared to 0.1%, P b 0.001) when compared with non-drug misusers. The drug misuse group also had higher rates of non-routine discharge Table 1 Demographics for Patients Undergoing Primary Total Hip or Total Knee Arthroplasty With Bivariate Analysis Comparing Those Who Misuse Drugs to Those Who Do Not Misuse Drugs. (SD, Standard Deviation). Parameter
Drug Misuse
No Drug Misuse
P Value
Number of patients, N Gender (%) Male Female Discharge dispostion (%) Routine/home Left AMA Non-routine Mortality Age, years, mean (SD) Days of care, mean (SD) Bed size b100 100–199 200–299 300–499 500+ Primary source of payment Private insurance Medicare Medicaid Workmen's Other
13,163
8,366,327
45.2 54.8
38.5 61.5
b0.001
48.7 0.4 38.8 1.4 51.6 (12.9) 5.3 (4.0)
51.8 0.1 30.0 0.3 67.3 (11.7) 5.1 (4.2)
b0.001
23.3 18.1 16.6 28.2 13.9
20.8 26.2 21.3 22.1 9.6
b0.001
31.4 35.3 18.9 6 8.4
31.9 59.6 2.9 1.0 4.6
0.217
b0.001 b0.001
M.J. Best et al. / The Journal of Arthroplasty 30 (2015) 1137–1141 Table 2 Breakdown of Drug Use by Type for Drug Misusers Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 13,163. Drug
Percentage (%)
Non-opioid mixed or unspecified Opioid Cannabis Cocaine Inhalants, absinthe, glue sniffing Sedative, hypnotic, anxiolytic Amphetamine Opioid and other drug mixed
47.3 26.1 10.9 10.2 6.5 4.8 2.7 1.2
Table 3 Prevalence of Comorbidities and Bivariate Analysis in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty Comparing Those Who Misuse Drugs to Those Who Do Not Misuse Drugs.
Comorbidity Number of patients, N Diabetes mellitus Hypertensive disease Previous myocardial infarction Atrial fibrillation Coronary artery disease Congestive heart failure Rheumatoid arthritis Osteoporosis
Drug Misuse Percentage
No Drug Misuse Percentage
P Value
13,163 15.3 25.7 1.7 0.8 0.4 2.5 4.1 1.0
8,366,327 13.0 46.9 2.1 4.2 4.9 2.7 3.2 3.1
b0.001 b0.001 b0.001 b0.001 b0.001 0.156 b0.001 b0.001
(38.8%) compared with non-drug misusers (30.0%) (P b 0.001). Medicare was the most common source of payment for both groups (Table 1). Analysis of the type of drug used showed that the most commonly used drugs were non-opioid mixed and unspecified combinations (47.3%), followed by opioids (26.1%), cannabis (10.9%) and cocaine (10.2%; Table 2). The drug misuse group had lower rates of hypertensive disease (25.7% compared to 46.9%), atrial fibrillation (0.8% compared to 4.2%), congestive heart failure (2.5% compared to 2.7%), previous myocardial infarction (1.7% compared to 2.1%), and coronary artery disease (0.4% compared to 4.9%) when compared with the group with no diagnosis of drug misuse (P b 0.001 for all; Table 3). Drug misusers had higher rates (P b 0.001) of surgery related complications including acute postoperative infection (0.7% compared to 0.2%) and acute postoperative anemia (25.8% compared to 17.3%) when compared with non-drug misusers (Fig. 1). Drug misusers also
1139
had higher rates (P b 0.001) of osteomyelitis (1.6% compared to 0.3%), cellulitis or abscess formation (0.4% compared to 0.2%), convulsion (3.4% compared to 0.4%), and blood transfusions (25.0% compared to 15.9%) than non-drug misusers. Infections of the joint prosthesis were higher (1.6% compared to 0.1%; P b 0.001) in the drug misusers group (Fig. 1). In multivariable logistic regression analysis, drug misuse had the second highest association with increased odds of having had a complication behind congestive heart failure (OR 2.035 range: 1.966–2.106, P b 0.001) (model fit: omnibus test of model coefficients: χ2 = 72,142, P b 0.001, Nagelkerke R2 = 0.014; Table 4). Multivariable logistic regression analysis showed that patients who misused drugs had an increased odds of having a prolonged hospital stay (OR 1.331 range: 1.284–1.381, P b 0.001) (omnibus χ2 = 72,142, P b 0.001, Nagelkerke R2 = 0.018; Table 5), an increased odds of leaving against medical advice (OR 7.742 range: 5.786–10.359, P b 0.001) (omnibus χ2 = 72,141, P b 0.001, Nagelkerke R2 = 0.033; Table 6) and an increased odds of mortality (OR 5.004 range: 4.316–5.803, P b 0.001) (model fit: omnibus test of model coefficients: χ2 = 72,141, P b 0.001, Nagelkerke R2 = 0.063; Table 7), which were comparable with that of the odds associated with patients with congestive heart failure and atrial fibrillation. Discussion This study describes the influence of drug misuse on perioperative outcomes following primary THA or TKA. Although individuals who misuse drugs are thought to be at increased risk of inflammatory arthropathies, the indication for surgery in this population is complicated by a higher risk of infection [7,8]. The results of this study show that patients who misuse drugs face prolonged hospital lengths of stay and higher odds of leaving against medical advice, mortality and complications following primary THA or TKA. In this study, the average length of hospital stay was longer in the group of patients with a diagnosis of drug misuse, alluding to possible increased healthcare utilization and healthcare cost in these patients. Although length of stay was only slightly longer in drug misusers, a more pronounced effect of drug misuse was seen when discharge status was analyzed. Drug misusers were nearly eight times more likely to leave against medical advice than patients who did not misuse drugs. In regression analysis, drug misuse was the most influential factor in leaving against medical advice. Numerous reports have found drug abuse to be associated with the decision to sign out against medical advice [22,23]. Reasons for this behavior have not been determined, but underlying addictive behaviors or the desire for more drugs are
Fig. 1. Prevalence of complications in patients undergoing primary total hip or total knee arthroplasty comparing those who misuse drugs to those who do not misuse drugs.
1140
M.J. Best et al. / The Journal of Arthroplasty 30 (2015) 1137–1141
Table 4 Multivariable Logistic Regression Analysis of Predictors of Perioperative Complications in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 8,379,490 (OR: Odds Ratio, CI: Confidence Interval). Parameter Congestive heart failure Drug misuse Age N80 Atrial fibrillation Osteoporosis Sex (F) Rheumatoid arthritis Age 65–80 Days of care Hospital bed size Race Primary source of payment Coronary artery disease Hypertensive disease Diabetes mellitus Age b50 Age 50–64
OR (95% CI)
P Value
Parameter
2.050 (2.033–2.067) 2.035 (1.966–2.106) 1.769 (1.761–1.776) 1.572 (1.562–1.583) 1.430 (1.419–1.442) 1.423 (1.418–1.427) 1.235 (1.225–1.245) 1.138 (1.135–1.142) 1.085 (1.084–1.086) 1.030 (1.009–1.052) 1.019 (1.011–1.026) 1.001 (0.998–1.004) 0.995 (0.988–1.001) 0.974 (0.971–0.976) 0.910 (0.916–0.914) 0.747 (0.743–0.752) 0.697 (0.695–0.699)
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.006 b0.001 0.424 0.117 b0.001 b0.001 b0.001 b0.001
Drug misuse Presence of complication Sex (F) Age 65–80 Hospital bed size Days of care Age N80 Race Atrial fibrillation Coronary artery disease Hypertensive disease Primary source of payment Age 50–64 Diabetes mellitus Age b50 Congestive heart failure Osteoporosis Rheumatoid arthritis
Omnibus χ2 = 72,142, P b 0.001. Nagelkerke R2 = 0.014.
Table 5 Multivariable Logistic Regression Analysis of Predictors of Prolonged Hospital Stay in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 8,379,490 (OR: Odds Ratio, CI: Confidence Interval).
Congestive heart failure Presence of complication Atrial fibrillation Age N80 Drug misuse Rheumatoid arthritis Age 65–80 Sex (F) Hospital bed size Primary source of payment Race Diabetes mellitus Age b50 Osteoporosis Age 50–64 Hypertensive disease Coronary artery disease Omnibus χ2 = 72,142, P b 0.001. Nagelkerke R2 = 0.018.
OR (95% CI)
P Value
7.742 (5.786–10.359) 2.641 (2.477–2.816) 2.224 (2.060–2.400) 1.806 (1.687–1.932) 1.193 (0.783–1.816) 1.016 (1.009–1.023) 0.972 (0.877–1.077) 0.953 (0.845–1.075) 0.837 (0.706–0.993) 0.821 (0.699–0.963) 0.789 (0.739–0.841) 0.768 (0.577–1.021) 0.597 (0.551–0.648) 0.459 (0.404–0.523) 0.374 (0.310–0.451) 0.009 (0.001–0.067) 0.008 (0.001–0.574) 0.008 (0.001–0.057)
b0.001 b0.001 b0.001 b0.001 0.411 b0.001 0.589 0.433 0.041 0.016 b0.001 0.069 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001
Omnibus χ2 = 72,141, P b 0.001. Nagelkerke R2 = 0.033.
proposed reasons [24–26]. Yu et al [11] also noted a loss to follow up in 27% of patients with a history of substance abuse, although in their analysis the majority of patients were alcohol abusers. This is the first study to demonstrate significantly increased odds of mortality among patients undergoing primary THA or TKA who misuse drugs. Although overall mortality for all patients in our study was low (0.3%), drug misusers were five times as likely to die in the perioperative period compared with patients with no diagnosis of drug misuse. Drug misuse was third only to congestive heart failure and atrial fibrillation as a prognostic indicator for mortality. There is a paucity of data on this topic and the few studies available involve heterogeneous patient populations, which may contribute to conflicting results. Hill et al [27] analyzed 40 patients who tested positive for cocaine and underwent various elective surgical procedures, the most common of which were orthopedic such as open reduction and internal fixation or hardware removal. They found no difference in morbidity and mortality when compared with non-drug users. Similarly, Ryb and Cooper [28] and Hadjizacharia et al [29] found no difference in mortality when comparing positive tested cocaine users and non-users undergoing surgery following trauma. There may be several explanations for these contrasting
Parameter
Table 6 Multivariable Logistic Regression Analysis of Predictors of Leaving Against Medical Advice in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 8,379,490 (OR: Odds Ratio, CI: Confidence Interval).
OR (95% CI)
P Value
2.780 (2.757–2.804) 1.717 (1.711–1.722) 1.653 (1.641–1.664) 1.514 (1.507–1.521) 1.331 (1.284–1.381) 1.317 (1.306–1.328) 1.310 (1.306–1.315) 1.155 (1.151–1.159) 1.083 (1.059–1.108) 1.007 (1.004–1.010) 0.955 (0.947–0.963) 0.820 (0.808–0.816) 0.734 (0.729–0.738) 0.675 (0.668–0.681) 0.626 (0.624–0.629) 0.560 (0.559–0.562) 0.534 (0.530–0.539)
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001
findings. The three mentioned studies collected urine samples on patients to screen for cocaine metabolites during their one time admission for elective surgery or following a traumatic injury. This may reflect transient or brief cocaine use and may not correlate with long-term use or a diagnosis of drug dependence. Additionally, it is possible these studies were underpowered to detect a difference, given the low mortality rate. In this study, drug misusers were twice as likely to have suffered a perioperative complication compared with those patients without a diagnosis of substance misuse. Many complications, such as postoperative anemia, blood transfusions, peripheral vascular complications, urinary tract infections, pulmonary insufficiency and convulsions were higher in the drug misusers group and have never been reported in this population in the perioperative period. Additionally, higher rates of infections were observed in this group, including postoperative infection, osteomyelitis, cellulitis or abscess formation, and infections of the joint prosthesis. Higher rates of infection were previously demonstrated in a study by Lehman et al [12], which evaluated 29 patients with human immunodeficiency virus or a history of intravenous drug use
Table 7 Multivariable Logistic Regression Analysis of Predictors of Mortality in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 8,379,490 (OR: Odds Ratio, CI: Confidence Interval). Parameter Congestive heart failure Atrial fibrillation Drug misuse Age N80 Presence of complication Diabetes mellitus Hospital bed size Sex (F) Days of care Primary source of payment Race Age 65–80 Rheumatoid arthritis Coronary artery disease Osteoporosis Hypertensive disease Age 50–64 Age b50 Omnibus χ2 = 72,141, P b 0.001. Nagelkerke R2 = 0.063.
OR (95% CI)
P Value
5.283 (5.081–5.492) 5.027 (4.861–5.199) 5.004 (4.316–5.803) 4.528 (4.407–4.652) 2.503 (2.439–2.569) 1.839 (1.782–1.898) 1.221 (0.961–1.551) 1.101 (1.072–1.131) 1.037 (0.989–1.088) 0.990 (0.972–1.009) 0.961 (0.896–1.032) 0.896 (0.873–0.919) 0.881 (0.814–0.952) 0.845 (0.793–0.901) 0.752 (0.691–0.819) 0.645 (0.627–0.662) 0.343 (0.330–0.357) 0.251 (0.229–0.275)
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.103 b0.001 b0.001 0.301 0.277 b0.001 b0.001 b0.001 b0.001 b 0.001 b 0.001 b 0.001
M.J. Best et al. / The Journal of Arthroplasty 30 (2015) 1137–1141
that underwent joint arthroplasty. In their analysis, both groups demonstrated increased rates of infection. In addition to joint prosthesis infection, the present study showed that other device complications were also higher in the drug misusers group. Wieser et al [7], retrospectively analyzed 27 primary THAs in patients with a history of drug use and found the five- and 10-year implant survival rates with failure for any reason were 61% and 52.3% and for septic reasons were 70.6% and 60.5%, respectively. The risk of infection is increased by the use of drugs [9,10] and the rate of periprosthetic joint infection after bacteremia, like that often seen in drug abusers, is reported to range from 30 to 40% [30]. This could explain the increased rates of infections and prosthesis complications in drug users in our study. Despite the strengths of using large, national databases for epidemiological research [31], this study has several limitations. First, the results are limited to practice patterns in the U.S. from 1990 to 2007, as more recent years could not be used due to high standard error from decreased sample size. Also, current drug use cannot be distinguished from past drug use in our patient population. While a remote history of drug use would likely have less of an effect on outcomes compared with active misuse, this was not evaluated in our study. Furthermore, identifying the influence of misuse of individual drugs was not possible due to limitations in sample size. The NHDS weighted sampling scheme has a standard error that increases exponentially as the sample size decreases. Consequently, the standard error is too high when each drug group is analyzed individually. Future prospective studies should seek to evaluate the influence of misuse of individual drugs on outcomes following primary THA or TKA to determine whether abuse of particular drugs confers a worse prognosis than others. Another limitation is that the database only allows for seven diagnosis codes and four procedure codes per entry. As a result, the prevalence of comorbid conditions and complications may be underreported [21]. Moreover, the severity of the comorbid disease cannot be appreciated when dichotomously classified [32]. Although mortality was a primary outcome analyzed in this study, the specific cause of death cannot be determined from the data. Another limitation is that the database only provides inpatient data, so complications that arise after discharge as well as follow-up data, are unknown. Similarly, the NHDS does not collect detailed information regarding anesthesia complications. Future research should evaluate whether drug misuse is associated with a higher rate of anesthesia complications during primary THA or TKA. In conclusion, this study is the largest investigation of the influence of substance abuse on perioperative outcomes following total joint arthroplasty in the U.S. Identifying risk factors associated with poor outcomes has the potential to change treatment strategies, resource allocation, in-hospital monitoring, and discharge planning for this patient population.
7.
8.
9.
10. 11.
12.
13. 14. 15.
16.
17.
18.
19. 20. 21.
22.
23. 24.
25. 26. 27. 28.
References 1. American Psychiatric Association (APA). Diagnostic and statistical manual of mental disorders. Text revision4th ed. . Washington, DC: APA; 2000. 2. Harwood HJ, Fountain D, Livermore G. Economic costs of alcohol abuse and alcoholism. Recent Dev Alcohol 1998;14:9751951. 3. McKusick D, Mark TL, King E, et al. Spending for mental health and substance abuse treatment, 1996. Health Aff (Millwood) 1998;17:59769578. 4. National Institute on Drug Abuse. Drug abuse and addiction: 25 years of discovery to advance the health of the people: the sixth triennial report to Congress from the secretary of Health and Human Services. Bethesda, MD: National Institute on Drug Abuse; 1999. 5. Office of National Drug Control Policy. The economic costs of drug abuse in the United States, 1992–1998. Washington, DC: Executive Office of the President; 2001. 6. Compton WM, Thomas YF, Stinson FS, et al. Prevalence, correlates, disability, and comorbidity of DSM-IV drug abuse and dependence in the United States: results from
29.
30.
31.
32.
1141
the national epidemiologic survey on alcohol and related conditions. Arch Gen Psychiatry 2007;64:566–76. Wieser K, Zingg PO, Betz M, et al. Total hip replacement in patients with history of illicit injecting drug use. Arch Orthop Trauma Surg 2012;132(7). http://dx.doi.org/10. 1007/s00402-012-1509-4 [22476213]. Lamas JL, Ocampo A, Martinez-Vazquez C, et al. Asymptomatic osteonecrosis of the hip in HIV-infected patients. Enferm Infecc Microbiol Clin 2010;28(3). http://dx.doi. org/10.1016/j.eimc.2009.06.003 [19683365]. Lowy FD, Miller M. New methods to investigate infectious disease transmission and pathogenesis–Staphylococcus aureus disease in drug users. Lancet Infect Dis 2002;2 [10S147330990200395X pii, 12383610]. Gordon RJ, Lowy FD. Bacterial infections in drug users. N Engl J Med 2005;353: 1945–54. Yu YH, Chen AC, Hu CC, et al. Acute delirium and poor compliance in total hip arthroplasty patients with substance abuse disorders. J Arthroplasty 2012;27(8). http://dx.doi.org/10.1016/j.arth.2011.12.003 [22325962]. Lehman CR, Ries MD, Paiement GD, et al. Infection after total joint arthroplasty in patients with human immunodeficiency virus or intravenous drug use. J Arthroplasty 2001;16 [3S0883-5403(01)68146-4, pii, 11307131]. Centers for Disease Control and Prevention. National Hospital Discharge Survey. http://www.cdc.gov/nchs/nhds.htm. [August 22, 2013; 2014:June 6]. Dennison C, Pokras R. Design and operation of the National Hospital Discharge Survey: 1988 redesign. Vital Health Stat 2000;1(39):3911261241. CDC/National Center for Health Statistics. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). http://www.cdc.gov/nchs/icd/ icd9cm.htm. [June 18, 2013; 2014:February/12]. Memtsoudis SG, Gonzalez Della Valle A, Besculides MC, et al. In-hospital complications and mortality of unilateral, bilateral, and revision TKA: based on an estimate of 4,159,661 discharges. Clin Orthop Relat Res 2008;466(11). http://dx.doi.org/10. 1007/s11999-008-0402-5 [18704616]. Stundner O, Kirksey M, Chiu YL, et al. Demographics and perioperative outcome in patients with depression and anxiety undergoing total joint arthroplasty: a population-based study. Psychosomatics 2013;54(2). http://dx.doi.org/10.1016/j. psym.2012.08.009 [23194934]. Hall MJOM. Comparison of the reliability of nationally representative estimates from the full (2007) and half (2008) sample National Hospital Discharge Surveys—implications for trend analyses. available from http://www.fcsm.gov/12papers/Hall_2012FCSM_VIII-B.pdf. [Accessed March, 1 2014]. Iezzoni LI, Daley J, Heeren T, et al. Using administrative data to screen hospitals for high complication rates. Inquiry 1994;31:18168908. Lemeshow S, Teres D, Klar J, et al. Mortality probability models (MPM II) based on an international cohort of intensive care unit patients. JAMA 1993;270:208230626. Bot AG, Menendez ME, Neuhaus V, et al. The influence of psychiatric comorbidity on perioperative outcomes after shoulder arthroplasty. J Shoulder Elbow Surg 2014; 23(4). http://dx.doi.org/10.1016/j.jse.2013.12.006 [24630546]. Pages KP, Russo JE, Wingerson DK, et al. Predictors and outcome of discharge against medical advice from the psychiatric units of a general hospital. Psychiatr Serv 1998; 49:99735960. Alfandre DJ. "I'm going home": discharges against medical advice. Mayo Clin Proc 2009;84(3). http://dx.doi.org/10.1016/S0025-6196(11)61143-9 [19252113]. Phillips DP, Christenfeld N, Ryan NM. An increase in the number of deaths in the United States in the first week of the month–an association with substance abuse and other causes of death. N Engl J Med 1999;341(2). http://dx.doi.org/10.1056/ NEJM199907083410206 [10395634]. Halpern SD, Mechem CC. Declining rate of substance abuse throughout the month. Am J Med 2001;110 [5S0002-9343(00)00749-X pii, 11286948]. Saitz R. Discharges against medical advice: time to address the causes. CMAJ 2002; 167:612358198. Hill GE, Ogunnaike BO, Johnson ER. General anaesthesia for the cocaine abusing patient. Is it safe? Br J Anaesth 2006;97 [5ael221, pii,16914461]. Ryb GE, Cooper C. Outcomes of cocaine-positive trauma patients undergoing surgery on the first day after admission. J Trauma 2008;65(4). http://dx.doi.org/10.1097/TA. 0b013e318187803f [18849795]. Hadjizacharia P, Green DJ, Plurad D, et al. Cocaine use in trauma: effect on injuries and outcomes. J Trauma 2009;66(2). http://dx.doi.org/10.1097/TA.0b013e3181622b9b [19204526]. Sendi P, Banderet F, Graber P, et al. Periprosthetic joint infection following Staphylococcus aureus bacteremia. J Infect 2011;63(1). http://dx.doi.org/10.1016/j.jinf.2011. 05.005 [21663971]. Bohl DD, Basques BA, Golinvaux NS, et al. Nationwide Inpatient Sample and National Surgical Quality Improvement Program give different results in hip fracture studies. Clin Orthop Relat Res 2014;472(6). http://dx.doi.org/10.1007/s11999-014-3559-0 [24615426]. Neuhaus V, Swellengrebel CH, Bossen JK, et al. What are the factors influencing outcome among patients admitted to a hospital with a proximal humeral fracture? Clin Orthop Relat Res 2013;471:5. http://dx.doi.org/10.1007/s11999-013-2876-z [23456187].
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Outcomes Following Primary Total Hip or Knee Arthroplasty in Substance Misusers Matthew J. Best, BS a,1, Leonard T. Buller, MD a,2, Alison K. Klika, MS b, Wael K. Barsoum, MD b,3 a b
University of Miami Miller School of Medicine, Department of Orthopaedic Surgery and Rehabilitation, Miami, Florida Cleveland Clinic Department of Orthopaedic Surgery, Mail Code A41,Cleveland, Ohio
a r t i c l e
i n f o
Article history: Received 20 November 2014 Accepted 31 January 2015 Keywords: total knee arthroplasty total hip arthroplasty drug misuse drug dependence drug abuse
a b s t r a c t The influence of drug misuse on outcomes following primary total hip (THA) or knee (TKA) arthroplasty is poorly understood. The National Hospital Discharge Survey was used to identify patients who underwent primary THA or TKA between 1990 and 2007. Patients were divided into two groups: 1) those with a diagnosis of drug misuse (cannabis, opioids, cocaine, amphetamines, sedatives, inhalants or mixed combinations) (n = 13,163) and 2) those with no diagnosis of misuse (n = 8,366,327). Patients with a diagnosis of drug misuse had longer hospital stays (P b 0.001), nearly eight times the odds of leaving against medical advice (P b 0.001) and five times the mortality rate (P b 0.001). Drug misuse was associated with higher odds (P b 0.001) of complications including postoperative infection, anemia, convulsions, osteomyelitis, and blood transfusion. © 2015 Elsevier Inc. All rights reserved.
Drug misuse, including substance abuse and dependence [1], is a maladaptive pattern of substance use with considerable societal, economic and personal cost [2–5]. The lifetime prevalence of drug abuse and dependence among U.S. adults is estimated at 7.7% and 2.6%, respectively [6]. The risk of secondary osteoarthritis among drug users is thought to be higher than the general population due to increased rates of osteonecrosis of the femoral head and inflammatory arthropathies due to repetitive bacteremia [7,8]. Additionally, when joint arthroplasty is considered, the risk of a septic complication is of concern as injection or inhalation of drugs is felt to increase the risk of infection [9,10]. Despite the rates of drug abuse and dependence in the U.S., few studies have evaluated their influence on outcomes following primary total hip (THA) or knee (TKA) arthroplasty [7,11,12]. Previous groups have found decreased implant survival rates [7], increased rates of infection [12], and high rates of postoperative substance withdrawal delirium [11] in substance abusers who underwent total joint arthroplasty, but these studies were limited by small sample size. The purpose of the present study was to measure the influence of drug misuse on inpatient perioperative outcomes following primary
One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2015.01.052. ⁎ Reprint requests: Leonard T. Buller, MD, Resident physician, University of Miami Department of Orthopaedic Surgery and Rehabilitation, Miami, FL. 1 Medical student, University of Miami Miller School of Medicine, Miami, FL, USA. 2 University of Miami Department of Orthopaedic Surgery and Rehabilitation, Miami, FL, USA. 3 Vice chair, Department of Orthopaedic Surgery, Cleveland Clinic, Cleveland, OH, USA. http://dx.doi.org/10.1016/j.arth.2015.01.052 0883-5403/© 2015 Elsevier Inc. All rights reserved.
THA or TKA using a large national database. The specific aims of the study were to evaluate the relationship between substance abuse and length of hospital stay, discharge disposition, mortality and perioperative complications in patients undergoing primary THA or TKA. The identification of modifiable risk factors associated with complications following total joint arthroplasty may allow surgeons to intervene preoperatively, potentially decreasing complications and improving outcomes. Materials and Methods Data Source The National Hospital Discharge Survey (NHDS) [13], developed by the National Center for Healthcare Statistics Division of the Centers for Disease Control and Prevention (CDC), was used in this study. The NHDS is considered the most comprehensive of all inpatient surgical databases in use today and is the principal database used by the U.S. government for monitoring hospital use [14]. Publicly available, the NHDS provides demographic and medical data for inpatients discharged from non-federal, short stay hospitals [14]. The survey uses International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9CM) [15] codes to classify up to seven discharge diagnoses and up to four procedures that are present at the time of discharge. In addition to medical information, the NHDS collects demographic information (age, gender), expected source of payment (insurance status), length of hospital stay, hospital size, U.S. region, and inpatient outcomes including discharge destination [16]. The NHDS ensures an unbiased national sampling by using a complex three-stage probability design
1138
M.J. Best et al. / The Journal of Arthroplasty 30 (2015) 1137–1141
including: inflation by reciprocals of the probabilities of sample selection, adjustment for no response and population weighting ratio adjustments [14]. This study did not require approval by the institutional review board because the NHDS is a publically available database with no patient identifying information.
Patient Selection All patients admitted to hospitals in the U.S. who underwent primary THA or TKA between 1990 and 2007 were identified using ICD-9-CM codes. Using previously described techniques, discharges with a procedure code (ICD-9-CM) of primary THA (81.51) or TKA (81.54) were identified [17]. Due to National Center for Health Statistics budgetary limitations starting in 2008, the number of hospital surveys was halved, decreasing the precision of the survey data and nearly doubling the relative standard error [18]. Consequently, we chose 2007 as the endpoint of our study. Patients were divided into two groups: 1) those with a diagnosis of drug misuse (ICD-9-CM: 304.00–304.93, 305.20–305.93) and 2) those who did not have a diagnosis of drug misuse. The drug misuse group includes subjects with diagnoses for either substance abuse or substance dependence. Substance abuse is defined as a maladaptive pattern of substance use leading to clinically significant impairment or distress as manifested by one or more of the following, occurring within a 12-month period: 1) recurrent substance use resulting in a failure to fulfill major role obligations at work, school, or home (e.g. repeated absences or poor work performance related to substance use; substancerelated absences, suspensions, or expulsions from school; neglect of children or household), 2) recurrent substance use in situations in which it is physically hazardous (e.g. driving an automobile or operating a machine when impaired) 3) recurrent substance-related legal problems (e.g. arrests for substance-related disorderly conduct) 4) continued substance use despite persistent or recurrent social or interpersonal problems caused or exacerbated by the effects of the substance (e.g. arguments with spouse about consequences of intoxication, physical fights) 5) the symptoms for substance abuse have never met the criteria for substance dependence [1]. In contrast, substance dependence is defined as a maladaptive pattern of substance use leading to clinically significant impairment or distress as manifested by three or more of the following, occurring at any time in the same 12-month period: 1) tolerance, as defined by either of the following: a) a need for markedly increased amounts of the substance to achieve intoxication or desired effect b) markedly diminished effect with continued use of the same amount of the substance, 2) withdrawal, as manifested by either of the following: a) the characteristic withdrawal syndrome for the substance, b) taking the same (or a closely related) substance to relieve or avoid withdrawal symptoms, 3) taking the substance often in larger amounts or over a longer period than was intended, 4) having a persistent desire or unsuccessful efforts to cut down or control substance use [1]. Demographic variables were collected including age, sex and prevalence of comorbidities. The length of hospital stay and discharge destination were determined. The incidence of complications was determined using the complication screening package [19]. The variable “perioperative complication” was created based upon the following (ICD-9-CM): acute postoperative bleeding (998.1), acute postoperative infection (998.5), other operative complication (998.89), acute postoperative anemia (285.1), thrombocytopenia (287.4, 287.5), peripheral vascular complication (997.2), urinary tract infection (599.0), other urinary complication (997.5), acute renal failure (584), acute myocardial infarction (410), pulmonary embolism (415.1), pulmonary insufficiency (518.5), acute deep venous thrombosis (453.4), osteomyelitis (730.0–730.2), cellulitis/abscess formation (682), convulsion (780.39), transfusion of blood (99.0), mechanical complication of internal orthopedic device (996.4, 996.79), infection of internal joint prosthesis (996.77), and other complication of internal joint prosthesis (996.77).
Statistical Analysis Differences between continuous variables were compared using the independent-samples t-test, while the Pearson chi square test was used to compare differences between categorical variables. To determine whether drug abuse was an independent predictor of a negative inhospital outcome, variables present in at least 2% of the population [20] were included in a multivariable binary logistic regression model. The dichotomous variables were 1) presence of any complication, 2) prolonged hospital stay (greater than the 75th percentile of the mean), 3) leaving against medical advice, and 4) in-hospital mortality. Potential confounders were controlled for using a multivariable regression model, to isolate the effect of drug abuse on inpatient outcomes. Covariates accounted for in the regression model included: gender, age, length of stay, hospital bed size, primary source of payment, presence of a complication, and preexisting comorbidities (i.e., diabetes mellitus, hypertension, congestive heart failure, coronary artery disease, atrial fibrillation, previous myocardial infarction, osteoporosis, and rheumatoid arthritis). Odds ratios and confidence intervals were calculated to assess the association between drug abuse and inpatient perioperative complications. Correcting for multiple comparisons, a Pvalue b0.001 was used to define statistical significance, as previously described [21]. All data were analyzed using the software-statistical package for social sciences [SPSS] version 20 (Chicago, IL, USA). Source of Funding No external funding source was used for the conduct of this study. Results A cohort representative of 8,379,490 patients who underwent primary THA or TKA between 1990 and 2007 was identified (Table 1). Of the total cohort, 13,163 patients had a diagnosis of drug misuse, while 8,366,327 patients had no diagnosis of drug misuse. The drug misuse group was younger (51.6 ± 12.9 years compared to 67.3 ± 11.7 years; P b 0.001), had longer hospital stays (5.3 ± 4.0 days compared to 5.1 ± 4.2 days; P b 0.001), and had a higher rate of patients who left against medical advice (0.4% compared to 0.1%, P b 0.001) when compared with non-drug misusers. The drug misuse group also had higher rates of non-routine discharge Table 1 Demographics for Patients Undergoing Primary Total Hip or Total Knee Arthroplasty With Bivariate Analysis Comparing Those Who Misuse Drugs to Those Who Do Not Misuse Drugs. (SD, Standard Deviation). Parameter
Drug Misuse
No Drug Misuse
P Value
Number of patients, N Gender (%) Male Female Discharge dispostion (%) Routine/home Left AMA Non-routine Mortality Age, years, mean (SD) Days of care, mean (SD) Bed size b100 100–199 200–299 300–499 500+ Primary source of payment Private insurance Medicare Medicaid Workmen's Other
13,163
8,366,327
45.2 54.8
38.5 61.5
b0.001
48.7 0.4 38.8 1.4 51.6 (12.9) 5.3 (4.0)
51.8 0.1 30.0 0.3 67.3 (11.7) 5.1 (4.2)
b0.001
23.3 18.1 16.6 28.2 13.9
20.8 26.2 21.3 22.1 9.6
b0.001
31.4 35.3 18.9 6 8.4
31.9 59.6 2.9 1.0 4.6
0.217
b0.001 b0.001
M.J. Best et al. / The Journal of Arthroplasty 30 (2015) 1137–1141 Table 2 Breakdown of Drug Use by Type for Drug Misusers Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 13,163. Drug
Percentage (%)
Non-opioid mixed or unspecified Opioid Cannabis Cocaine Inhalants, absinthe, glue sniffing Sedative, hypnotic, anxiolytic Amphetamine Opioid and other drug mixed
47.3 26.1 10.9 10.2 6.5 4.8 2.7 1.2
Table 3 Prevalence of Comorbidities and Bivariate Analysis in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty Comparing Those Who Misuse Drugs to Those Who Do Not Misuse Drugs.
Comorbidity Number of patients, N Diabetes mellitus Hypertensive disease Previous myocardial infarction Atrial fibrillation Coronary artery disease Congestive heart failure Rheumatoid arthritis Osteoporosis
Drug Misuse Percentage
No Drug Misuse Percentage
P Value
13,163 15.3 25.7 1.7 0.8 0.4 2.5 4.1 1.0
8,366,327 13.0 46.9 2.1 4.2 4.9 2.7 3.2 3.1
b0.001 b0.001 b0.001 b0.001 b0.001 0.156 b0.001 b0.001
(38.8%) compared with non-drug misusers (30.0%) (P b 0.001). Medicare was the most common source of payment for both groups (Table 1). Analysis of the type of drug used showed that the most commonly used drugs were non-opioid mixed and unspecified combinations (47.3%), followed by opioids (26.1%), cannabis (10.9%) and cocaine (10.2%; Table 2). The drug misuse group had lower rates of hypertensive disease (25.7% compared to 46.9%), atrial fibrillation (0.8% compared to 4.2%), congestive heart failure (2.5% compared to 2.7%), previous myocardial infarction (1.7% compared to 2.1%), and coronary artery disease (0.4% compared to 4.9%) when compared with the group with no diagnosis of drug misuse (P b 0.001 for all; Table 3). Drug misusers had higher rates (P b 0.001) of surgery related complications including acute postoperative infection (0.7% compared to 0.2%) and acute postoperative anemia (25.8% compared to 17.3%) when compared with non-drug misusers (Fig. 1). Drug misusers also
1139
had higher rates (P b 0.001) of osteomyelitis (1.6% compared to 0.3%), cellulitis or abscess formation (0.4% compared to 0.2%), convulsion (3.4% compared to 0.4%), and blood transfusions (25.0% compared to 15.9%) than non-drug misusers. Infections of the joint prosthesis were higher (1.6% compared to 0.1%; P b 0.001) in the drug misusers group (Fig. 1). In multivariable logistic regression analysis, drug misuse had the second highest association with increased odds of having had a complication behind congestive heart failure (OR 2.035 range: 1.966–2.106, P b 0.001) (model fit: omnibus test of model coefficients: χ2 = 72,142, P b 0.001, Nagelkerke R2 = 0.014; Table 4). Multivariable logistic regression analysis showed that patients who misused drugs had an increased odds of having a prolonged hospital stay (OR 1.331 range: 1.284–1.381, P b 0.001) (omnibus χ2 = 72,142, P b 0.001, Nagelkerke R2 = 0.018; Table 5), an increased odds of leaving against medical advice (OR 7.742 range: 5.786–10.359, P b 0.001) (omnibus χ2 = 72,141, P b 0.001, Nagelkerke R2 = 0.033; Table 6) and an increased odds of mortality (OR 5.004 range: 4.316–5.803, P b 0.001) (model fit: omnibus test of model coefficients: χ2 = 72,141, P b 0.001, Nagelkerke R2 = 0.063; Table 7), which were comparable with that of the odds associated with patients with congestive heart failure and atrial fibrillation. Discussion This study describes the influence of drug misuse on perioperative outcomes following primary THA or TKA. Although individuals who misuse drugs are thought to be at increased risk of inflammatory arthropathies, the indication for surgery in this population is complicated by a higher risk of infection [7,8]. The results of this study show that patients who misuse drugs face prolonged hospital lengths of stay and higher odds of leaving against medical advice, mortality and complications following primary THA or TKA. In this study, the average length of hospital stay was longer in the group of patients with a diagnosis of drug misuse, alluding to possible increased healthcare utilization and healthcare cost in these patients. Although length of stay was only slightly longer in drug misusers, a more pronounced effect of drug misuse was seen when discharge status was analyzed. Drug misusers were nearly eight times more likely to leave against medical advice than patients who did not misuse drugs. In regression analysis, drug misuse was the most influential factor in leaving against medical advice. Numerous reports have found drug abuse to be associated with the decision to sign out against medical advice [22,23]. Reasons for this behavior have not been determined, but underlying addictive behaviors or the desire for more drugs are
Fig. 1. Prevalence of complications in patients undergoing primary total hip or total knee arthroplasty comparing those who misuse drugs to those who do not misuse drugs.
1140
M.J. Best et al. / The Journal of Arthroplasty 30 (2015) 1137–1141
Table 4 Multivariable Logistic Regression Analysis of Predictors of Perioperative Complications in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 8,379,490 (OR: Odds Ratio, CI: Confidence Interval). Parameter Congestive heart failure Drug misuse Age N80 Atrial fibrillation Osteoporosis Sex (F) Rheumatoid arthritis Age 65–80 Days of care Hospital bed size Race Primary source of payment Coronary artery disease Hypertensive disease Diabetes mellitus Age b50 Age 50–64
OR (95% CI)
P Value
Parameter
2.050 (2.033–2.067) 2.035 (1.966–2.106) 1.769 (1.761–1.776) 1.572 (1.562–1.583) 1.430 (1.419–1.442) 1.423 (1.418–1.427) 1.235 (1.225–1.245) 1.138 (1.135–1.142) 1.085 (1.084–1.086) 1.030 (1.009–1.052) 1.019 (1.011–1.026) 1.001 (0.998–1.004) 0.995 (0.988–1.001) 0.974 (0.971–0.976) 0.910 (0.916–0.914) 0.747 (0.743–0.752) 0.697 (0.695–0.699)
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.006 b0.001 0.424 0.117 b0.001 b0.001 b0.001 b0.001
Drug misuse Presence of complication Sex (F) Age 65–80 Hospital bed size Days of care Age N80 Race Atrial fibrillation Coronary artery disease Hypertensive disease Primary source of payment Age 50–64 Diabetes mellitus Age b50 Congestive heart failure Osteoporosis Rheumatoid arthritis
Omnibus χ2 = 72,142, P b 0.001. Nagelkerke R2 = 0.014.
Table 5 Multivariable Logistic Regression Analysis of Predictors of Prolonged Hospital Stay in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 8,379,490 (OR: Odds Ratio, CI: Confidence Interval).
Congestive heart failure Presence of complication Atrial fibrillation Age N80 Drug misuse Rheumatoid arthritis Age 65–80 Sex (F) Hospital bed size Primary source of payment Race Diabetes mellitus Age b50 Osteoporosis Age 50–64 Hypertensive disease Coronary artery disease Omnibus χ2 = 72,142, P b 0.001. Nagelkerke R2 = 0.018.
OR (95% CI)
P Value
7.742 (5.786–10.359) 2.641 (2.477–2.816) 2.224 (2.060–2.400) 1.806 (1.687–1.932) 1.193 (0.783–1.816) 1.016 (1.009–1.023) 0.972 (0.877–1.077) 0.953 (0.845–1.075) 0.837 (0.706–0.993) 0.821 (0.699–0.963) 0.789 (0.739–0.841) 0.768 (0.577–1.021) 0.597 (0.551–0.648) 0.459 (0.404–0.523) 0.374 (0.310–0.451) 0.009 (0.001–0.067) 0.008 (0.001–0.574) 0.008 (0.001–0.057)
b0.001 b0.001 b0.001 b0.001 0.411 b0.001 0.589 0.433 0.041 0.016 b0.001 0.069 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001
Omnibus χ2 = 72,141, P b 0.001. Nagelkerke R2 = 0.033.
proposed reasons [24–26]. Yu et al [11] also noted a loss to follow up in 27% of patients with a history of substance abuse, although in their analysis the majority of patients were alcohol abusers. This is the first study to demonstrate significantly increased odds of mortality among patients undergoing primary THA or TKA who misuse drugs. Although overall mortality for all patients in our study was low (0.3%), drug misusers were five times as likely to die in the perioperative period compared with patients with no diagnosis of drug misuse. Drug misuse was third only to congestive heart failure and atrial fibrillation as a prognostic indicator for mortality. There is a paucity of data on this topic and the few studies available involve heterogeneous patient populations, which may contribute to conflicting results. Hill et al [27] analyzed 40 patients who tested positive for cocaine and underwent various elective surgical procedures, the most common of which were orthopedic such as open reduction and internal fixation or hardware removal. They found no difference in morbidity and mortality when compared with non-drug users. Similarly, Ryb and Cooper [28] and Hadjizacharia et al [29] found no difference in mortality when comparing positive tested cocaine users and non-users undergoing surgery following trauma. There may be several explanations for these contrasting
Parameter
Table 6 Multivariable Logistic Regression Analysis of Predictors of Leaving Against Medical Advice in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 8,379,490 (OR: Odds Ratio, CI: Confidence Interval).
OR (95% CI)
P Value
2.780 (2.757–2.804) 1.717 (1.711–1.722) 1.653 (1.641–1.664) 1.514 (1.507–1.521) 1.331 (1.284–1.381) 1.317 (1.306–1.328) 1.310 (1.306–1.315) 1.155 (1.151–1.159) 1.083 (1.059–1.108) 1.007 (1.004–1.010) 0.955 (0.947–0.963) 0.820 (0.808–0.816) 0.734 (0.729–0.738) 0.675 (0.668–0.681) 0.626 (0.624–0.629) 0.560 (0.559–0.562) 0.534 (0.530–0.539)
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001
findings. The three mentioned studies collected urine samples on patients to screen for cocaine metabolites during their one time admission for elective surgery or following a traumatic injury. This may reflect transient or brief cocaine use and may not correlate with long-term use or a diagnosis of drug dependence. Additionally, it is possible these studies were underpowered to detect a difference, given the low mortality rate. In this study, drug misusers were twice as likely to have suffered a perioperative complication compared with those patients without a diagnosis of substance misuse. Many complications, such as postoperative anemia, blood transfusions, peripheral vascular complications, urinary tract infections, pulmonary insufficiency and convulsions were higher in the drug misusers group and have never been reported in this population in the perioperative period. Additionally, higher rates of infections were observed in this group, including postoperative infection, osteomyelitis, cellulitis or abscess formation, and infections of the joint prosthesis. Higher rates of infection were previously demonstrated in a study by Lehman et al [12], which evaluated 29 patients with human immunodeficiency virus or a history of intravenous drug use
Table 7 Multivariable Logistic Regression Analysis of Predictors of Mortality in Patients Undergoing Primary Total Hip or Total Knee Arthroplasty, N = 8,379,490 (OR: Odds Ratio, CI: Confidence Interval). Parameter Congestive heart failure Atrial fibrillation Drug misuse Age N80 Presence of complication Diabetes mellitus Hospital bed size Sex (F) Days of care Primary source of payment Race Age 65–80 Rheumatoid arthritis Coronary artery disease Osteoporosis Hypertensive disease Age 50–64 Age b50 Omnibus χ2 = 72,141, P b 0.001. Nagelkerke R2 = 0.063.
OR (95% CI)
P Value
5.283 (5.081–5.492) 5.027 (4.861–5.199) 5.004 (4.316–5.803) 4.528 (4.407–4.652) 2.503 (2.439–2.569) 1.839 (1.782–1.898) 1.221 (0.961–1.551) 1.101 (1.072–1.131) 1.037 (0.989–1.088) 0.990 (0.972–1.009) 0.961 (0.896–1.032) 0.896 (0.873–0.919) 0.881 (0.814–0.952) 0.845 (0.793–0.901) 0.752 (0.691–0.819) 0.645 (0.627–0.662) 0.343 (0.330–0.357) 0.251 (0.229–0.275)
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.103 b0.001 b0.001 0.301 0.277 b0.001 b0.001 b0.001 b0.001 b 0.001 b 0.001 b 0.001
M.J. Best et al. / The Journal of Arthroplasty 30 (2015) 1137–1141
that underwent joint arthroplasty. In their analysis, both groups demonstrated increased rates of infection. In addition to joint prosthesis infection, the present study showed that other device complications were also higher in the drug misusers group. Wieser et al [7], retrospectively analyzed 27 primary THAs in patients with a history of drug use and found the five- and 10-year implant survival rates with failure for any reason were 61% and 52.3% and for septic reasons were 70.6% and 60.5%, respectively. The risk of infection is increased by the use of drugs [9,10] and the rate of periprosthetic joint infection after bacteremia, like that often seen in drug abusers, is reported to range from 30 to 40% [30]. This could explain the increased rates of infections and prosthesis complications in drug users in our study. Despite the strengths of using large, national databases for epidemiological research [31], this study has several limitations. First, the results are limited to practice patterns in the U.S. from 1990 to 2007, as more recent years could not be used due to high standard error from decreased sample size. Also, current drug use cannot be distinguished from past drug use in our patient population. While a remote history of drug use would likely have less of an effect on outcomes compared with active misuse, this was not evaluated in our study. Furthermore, identifying the influence of misuse of individual drugs was not possible due to limitations in sample size. The NHDS weighted sampling scheme has a standard error that increases exponentially as the sample size decreases. Consequently, the standard error is too high when each drug group is analyzed individually. Future prospective studies should seek to evaluate the influence of misuse of individual drugs on outcomes following primary THA or TKA to determine whether abuse of particular drugs confers a worse prognosis than others. Another limitation is that the database only allows for seven diagnosis codes and four procedure codes per entry. As a result, the prevalence of comorbid conditions and complications may be underreported [21]. Moreover, the severity of the comorbid disease cannot be appreciated when dichotomously classified [32]. Although mortality was a primary outcome analyzed in this study, the specific cause of death cannot be determined from the data. Another limitation is that the database only provides inpatient data, so complications that arise after discharge as well as follow-up data, are unknown. Similarly, the NHDS does not collect detailed information regarding anesthesia complications. Future research should evaluate whether drug misuse is associated with a higher rate of anesthesia complications during primary THA or TKA. In conclusion, this study is the largest investigation of the influence of substance abuse on perioperative outcomes following total joint arthroplasty in the U.S. Identifying risk factors associated with poor outcomes has the potential to change treatment strategies, resource allocation, in-hospital monitoring, and discharge planning for this patient population.
7.
8.
9.
10. 11.
12.
13. 14. 15.
16.
17.
18.
19. 20. 21.
22.
23. 24.
25. 26. 27. 28.
References 1. American Psychiatric Association (APA). Diagnostic and statistical manual of mental disorders. Text revision4th ed. . Washington, DC: APA; 2000. 2. Harwood HJ, Fountain D, Livermore G. Economic costs of alcohol abuse and alcoholism. Recent Dev Alcohol 1998;14:9751951. 3. McKusick D, Mark TL, King E, et al. Spending for mental health and substance abuse treatment, 1996. Health Aff (Millwood) 1998;17:59769578. 4. National Institute on Drug Abuse. Drug abuse and addiction: 25 years of discovery to advance the health of the people: the sixth triennial report to Congress from the secretary of Health and Human Services. Bethesda, MD: National Institute on Drug Abuse; 1999. 5. Office of National Drug Control Policy. The economic costs of drug abuse in the United States, 1992–1998. Washington, DC: Executive Office of the President; 2001. 6. Compton WM, Thomas YF, Stinson FS, et al. Prevalence, correlates, disability, and comorbidity of DSM-IV drug abuse and dependence in the United States: results from
29.
30.
31.
32.
1141
the national epidemiologic survey on alcohol and related conditions. Arch Gen Psychiatry 2007;64:566–76. Wieser K, Zingg PO, Betz M, et al. Total hip replacement in patients with history of illicit injecting drug use. Arch Orthop Trauma Surg 2012;132(7). http://dx.doi.org/10. 1007/s00402-012-1509-4 [22476213]. Lamas JL, Ocampo A, Martinez-Vazquez C, et al. Asymptomatic osteonecrosis of the hip in HIV-infected patients. Enferm Infecc Microbiol Clin 2010;28(3). http://dx.doi. org/10.1016/j.eimc.2009.06.003 [19683365]. Lowy FD, Miller M. New methods to investigate infectious disease transmission and pathogenesis–Staphylococcus aureus disease in drug users. Lancet Infect Dis 2002;2 [10S147330990200395X pii, 12383610]. Gordon RJ, Lowy FD. Bacterial infections in drug users. N Engl J Med 2005;353: 1945–54. Yu YH, Chen AC, Hu CC, et al. Acute delirium and poor compliance in total hip arthroplasty patients with substance abuse disorders. J Arthroplasty 2012;27(8). http://dx.doi.org/10.1016/j.arth.2011.12.003 [22325962]. Lehman CR, Ries MD, Paiement GD, et al. Infection after total joint arthroplasty in patients with human immunodeficiency virus or intravenous drug use. J Arthroplasty 2001;16 [3S0883-5403(01)68146-4, pii, 11307131]. Centers for Disease Control and Prevention. National Hospital Discharge Survey. http://www.cdc.gov/nchs/nhds.htm. [August 22, 2013; 2014:June 6]. Dennison C, Pokras R. Design and operation of the National Hospital Discharge Survey: 1988 redesign. Vital Health Stat 2000;1(39):3911261241. CDC/National Center for Health Statistics. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). http://www.cdc.gov/nchs/icd/ icd9cm.htm. [June 18, 2013; 2014:February/12]. Memtsoudis SG, Gonzalez Della Valle A, Besculides MC, et al. In-hospital complications and mortality of unilateral, bilateral, and revision TKA: based on an estimate of 4,159,661 discharges. Clin Orthop Relat Res 2008;466(11). http://dx.doi.org/10. 1007/s11999-008-0402-5 [18704616]. Stundner O, Kirksey M, Chiu YL, et al. Demographics and perioperative outcome in patients with depression and anxiety undergoing total joint arthroplasty: a population-based study. Psychosomatics 2013;54(2). http://dx.doi.org/10.1016/j. psym.2012.08.009 [23194934]. Hall MJOM. Comparison of the reliability of nationally representative estimates from the full (2007) and half (2008) sample National Hospital Discharge Surveys—implications for trend analyses. available from http://www.fcsm.gov/12papers/Hall_2012FCSM_VIII-B.pdf. [Accessed March, 1 2014]. Iezzoni LI, Daley J, Heeren T, et al. Using administrative data to screen hospitals for high complication rates. Inquiry 1994;31:18168908. Lemeshow S, Teres D, Klar J, et al. Mortality probability models (MPM II) based on an international cohort of intensive care unit patients. JAMA 1993;270:208230626. Bot AG, Menendez ME, Neuhaus V, et al. The influence of psychiatric comorbidity on perioperative outcomes after shoulder arthroplasty. J Shoulder Elbow Surg 2014; 23(4). http://dx.doi.org/10.1016/j.jse.2013.12.006 [24630546]. Pages KP, Russo JE, Wingerson DK, et al. Predictors and outcome of discharge against medical advice from the psychiatric units of a general hospital. Psychiatr Serv 1998; 49:99735960. Alfandre DJ. "I'm going home": discharges against medical advice. Mayo Clin Proc 2009;84(3). http://dx.doi.org/10.1016/S0025-6196(11)61143-9 [19252113]. Phillips DP, Christenfeld N, Ryan NM. An increase in the number of deaths in the United States in the first week of the month–an association with substance abuse and other causes of death. N Engl J Med 1999;341(2). http://dx.doi.org/10.1056/ NEJM199907083410206 [10395634]. Halpern SD, Mechem CC. Declining rate of substance abuse throughout the month. Am J Med 2001;110 [5S0002-9343(00)00749-X pii, 11286948]. Saitz R. Discharges against medical advice: time to address the causes. CMAJ 2002; 167:612358198. Hill GE, Ogunnaike BO, Johnson ER. General anaesthesia for the cocaine abusing patient. Is it safe? Br J Anaesth 2006;97 [5ael221, pii,16914461]. Ryb GE, Cooper C. Outcomes of cocaine-positive trauma patients undergoing surgery on the first day after admission. J Trauma 2008;65(4). http://dx.doi.org/10.1097/TA. 0b013e318187803f [18849795]. Hadjizacharia P, Green DJ, Plurad D, et al. Cocaine use in trauma: effect on injuries and outcomes. J Trauma 2009;66(2). http://dx.doi.org/10.1097/TA.0b013e3181622b9b [19204526]. Sendi P, Banderet F, Graber P, et al. Periprosthetic joint infection following Staphylococcus aureus bacteremia. J Infect 2011;63(1). http://dx.doi.org/10.1016/j.jinf.2011. 05.005 [21663971]. Bohl DD, Basques BA, Golinvaux NS, et al. Nationwide Inpatient Sample and National Surgical Quality Improvement Program give different results in hip fracture studies. Clin Orthop Relat Res 2014;472(6). http://dx.doi.org/10.1007/s11999-014-3559-0 [24615426]. Neuhaus V, Swellengrebel CH, Bossen JK, et al. What are the factors influencing outcome among patients admitted to a hospital with a proximal humeral fracture? Clin Orthop Relat Res 2013;471:5. http://dx.doi.org/10.1007/s11999-013-2876-z [23456187].
