Authors: Zachary Ramsey, Victoria Ierulli, Nicholas O. Gerard, III, Shreya M. Saraf, Wendell Heard, Mary K. Mulcahey
Categories: Original Article
Source: Arthroscopy, Sports Medicine, and Rehabilitation
Authors: Zachary Ramsey, Victoria Ierulli, Nicholas O. Gerard, Shreya M. Saraf, Wendell Heard, Mary K. Mulcahey
To estimate the prevalence of psychotropic medication (PTM) usage in patients undergoing knee arthroscopy and to identify knee pathologies associated with the use of PTMs.
A retrospective chart review was conducted. Subjects were identified using the billing database at a single institution. All eligible patients treated by 2 fellowship-trained orthopaedic sports medicine surgeons between 2017 and 2021 were analyzed; patients were included if they were at least 12 years old and had at least 6 months of documented follow-up. The following characteristics were age, sex, surgery laterality, ipsilateral knee surgery history, knee pathology, and use of PTMs if listed in patient charts within 6 months prior to surgery.
Records from 1,000 arthroscopic knee surgical procedures were examined (399 female and 601 male patients). From 2017 to 2021, 245 patients (24.5%) were prescribed at least 1 psychotropic drug. Female patients took PTM drugs at higher rates than male patients (odds ratio, 1.92; 95% confidence interval, 1.42-2.60; P < .001). An examination of drug classes revealed 106 antidepressant prescriptions (35.2%), 77 sedative hypnotics (25.6%), 38 stimulants (12.6%), 6 antipsychotics (2.0%), and 74 anticonvulsants/mood stabilizers (24.6%). Patients taking PTMs had lower adjusted odds of presenting with anterior cruciate ligament injury compared with their peers not taking PTMs (adjusted odds ratio, 0.55; 95% confidence interval, 0.32-0.82; P = .003).
The rate of PTM usage among patients undergoing knee arthroscopy was 24.5% based on 5 years of records from 1 institution. Female patients had a higher frequency of PTM use compared with their male counterparts.
PTM use is common among knee arthroscopy patients, with higher rates observed in female patients.
Psychotropic medication (PTM) usage has steadily increased in the United States since the 1990s.^1^^,^^2^ In 2007, researchers at the Centers for Disease Control and Prevention reported that 10% of the adult population in the United States was taking at least 1 PTM.^3^ Over a decade later, the National Health Interview Survey found that this rate rose to 15.8% in 2019 and 16.5% in 2020.^4^^,^^5^ This upward trajectory has been exacerbated by global events, most notably the COVID-19 pandemic. Time-series analysis of PTM prevalence has shown increased rates after the pandemic.^6^
Despite the high frequency of PTM use among Americans, there is a paucity of literature regarding the rate of PTM usage among patients undergoing knee arthroscopy and the impact of taking these medications on postoperative outcomes. Identifying PTM usage is important for improving patient care not only because psychopathologies such as depression and post-traumatic stress disorder have been associated with higher pain levels at discharge, but also because PTMs may increase perioperative risk and worsen postoperative outcomes.7, 8, 9 Specifically, anxiety and depression have been shown to increase the intensity of surgical pain, the risk of chronic postsurgical pain, and the risk of delirium.^10^^,^^11^ A retrospective analysis of patients undergoing rotator cuff repair, conducted using the TriNetX database (Cambridge, MA), further supported these Antidepressant use is associated with higher odds of acute postoperative pain, emergency department visits, and revision surgery.^12^ Physiological mechanisms may explain these findings as medications such as tricyclic antidepressants, selective serotonin reuptake inhibitors, and monoamine oxide inhibitors interact with anesthesia, with effects ranging from abnormal bleeding to hypertensive crisis and arrhythmia.^9^^,^^13^ Postoperatively, orthopaedic patients with mental health comorbidities also face higher rates of non-routine discharge, increased postoperative care costs, and higher rates of mortality.^14^^,^^15^
By means of retrospective health record review, previous studies have found that 48% of patients undergoing hip arthroscopy and 41.8% of patients undergoing shoulder arthroscopy use PTMs, as compared with a general population usage rate of 16.5%.^5^^,^^8^^,^^16^ The purpose of this study was to estimate the prevalence of PTM usage in patients undergoing knee arthroscopy and to identify knee pathologies associated with the use of PTMs. We hypothesized that the proportion of patients undergoing knee arthroscopy who were taking PTMs prior to surgery would be comparable to the proportions reported in patients undergoing shoulder and hip arthroscopy.
A retrospective chart review was conducted using the billing database at a single institution after expedited approval was received from the institutional review board of the overseeing institution. Patients treated by 2 fellowship-trained sports medicine surgeons (M.K.M., W.H.) between 2017 and 2021 were analyzed. Patients were included if they met the following age of at least 12 years, any type of arthroscopic knee procedure, and at least 6 months of documented postoperative follow-up. Variables including year of surgery, age, sex, laterality of procedure, history of ipsilateral knee surgery, PTM usage, and knee pathology were collected from patient charts. Patients were considered to be taking PTMs if the medications were listed as active in the electronic medical record at any time 6 months prior to surgery.
PTMs were classified as (1) antidepressants/anxiolytics, (2) sedative hypnotics, (3) stimulants, (4) antipsychotics, and (5) anticonvulsants/mood stabilizers. Table 1 lists the 5 drug classes and individual drugs within each class. For patients taking multiple PTMs, each drug was recorded and the presence of multiple medications was noted. All data analysis was conducted using R statistical software, version 4.3.2 (The R Foundation for Statistical Computing, Vienna, Austria).Table 1Drug Classes and Individual Drugs Seen in Patient ChartsDrug ClassSubtypesAntidepressants/anxiolyticsSSRIs: sertraline, fluoxetine, citalopram, escitalopram, paroxetine, fluvoxamineSNRIs: duloxetine, venlafaxineTCAs: amitriptyline, doxepin, nortriptylineAtypical bupropion, trazodone, vilazodone, mirtazapineSedative hypnoticsBenzodiazepines: alprazolam, diazepam, lorazepam, clonazepam, triazolamGABAergic topiramate, zolpidem, eszopicloneBarbiturates: butalbital, butabarbitalStimulantsAmphetamines: mixed-amphetamine salts, lisdexamfetamine, amphetamineOther: dexmethylphenidate, methylphenidateAntipsychoticsAtypical aripiprazole, lurasidoneDopamine ropinirolePhenothiazine prochlorperazineAnticonvulsants/mood stabilizersAnticonvulsants: pregabalin, gabapentinAntiepileptics: lamotrigine, oxcarbazepine, zonisamide, rufinamideMuscle tizanidine, cyclobenzaprineMood lithium, divalproex sodiumGABAergic, affective gamma-aminobutyric acid transmission; SNRI, selective norepinephrine reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor; TCA, tricyclic antidepressant.
Patient charts from 2017 to 2021 included 1,000 individual arthroscopic knee surgical procedures performed on 399 female patients (39.9%) and 601 male patients (60.1%). During this 5-year period, a total of 245 patients (24.5%) were taking at least 1 psychotropic drug, with 44 (4.4%) taking 2 or more drugs. In 2017, 43 patients (22.5%) were taking PTMs, compared with 57 (24.8%) in 2018, 59 (24.9%) in 2019, 39 (24.7%) in 2020, and 47 (25.5%) in 2021 (Fig 1). Although there was an empirical increase in drug usage rates from 2017 to 2021, year-on-year differences in drug usage rates were not statistically significant per the Mann-Kendall test (P = .22).Fig 1Psychotropic medication (PTM) usage in patients undergoing knee arthroscopy from 2017 to 2021.
From 2018 to 2021, 20 patients (10.5%) underwent repeated knee 5 (13.2%) were taking PTMs, and 15 (11.3%) were not. Logistic regression, holding all other variables constant and using 2017 as the base year, revealed that PTM status in a patient undergoing knee arthroscopy was not significantly associated with the probability of repeated arthroscopy for 4 years after surgery (P = .801).
A total of 128 female patients (32.1%) were found to use PTMs compared with 117 male patients (19.5%) (Fig 2). The odds ratio was calculated to be 1.92 (95% confidence interval [CI], 1.42-2.60; P < .001) using logistic regression. Other variables including history of ipsilateral knee surgery (P = .26) and laterality of surgery (P = .33) were not found to be statistically associated with psychotropic drug usage. Besides patient sex, the only variable associated with an increased probability of PTM usage was age (P < .001). Calculating the marginal effect of age, we observed that an increase in age by 10 years from the mean (34.1 years) was associated with an increase in the probability of PTM usage by 5.9%, holding all other variables constant at baseline.Fig 2Combined psychotropic medication (PTM) usage rates and subclass usage in all patients undergoing knee arthroscopy (A), male patients only (B), and female patients only (C) from 2017 to 2021.
Of the 301 prescribed medications, 106 (35.2%) were antidepressants, 77 (25.6%) were sedative hypnotics, 74 (24.6%) were anticonvulsants/mood stabilizers, 38 (12.6%) were stimulants, and 6 (2.0%) were antipsychotics. Trends in psychotropic usage by class were analyzed from 2017 to 2021 using Mann-Kendall tests. The tests failed to provide evidence of significant time trends within any of the drug classes (P > .8 across all classes), indicating that there was no discernible upward or downward trend in drug usage for each of the psychotropic classes (Fig 1). Of the 601 male patients, 41 (6.8%) were taking antidepressants, 33 (5.5%) were taking anticonvulsants/mood stabilizers, 32 (5.3%) were taking sedatives, 28 (4.7%) were taking stimulants, and 1 (0.2%) was taking antipsychotics. Of the 399 female patients, 58 (14.5%) were taking antidepressants, 41 (10.2%) were taking anticonvulsants/mood stabilizers, 39 (9.8%) were taking sedatives, 10 (2.5%) were taking stimulants, and 5 (1.3%) were taking antipsychotics. Within the subpopulation of patients taking PTMs, female patients had a higher frequency of PTM usage than male patients across every drug class except stimulants, with 7.8% for female patients versus 23.9% for male patients (antidepressants, 45.3% for female patients vs 35.0% for male patients; sedatives, 30.5% vs 27.4%; antipsychotics, 5.5% vs 0.9%; and anticonvulsants/mood stabilizers, 32.0% vs 28.2%). Male versus female PTM usage patterns by class are summarized in Figure 2. These data indicate that female patients were more frequently prescribed PTMs from multiple classes when compared with their male counterparts. Within the patient population taking PTMs, 9 female patients (7.0%) and 4 male patients (3.4%) took multiple concurrent medications from different psychotropic classes. The most prescribed medications by class were sertraline (antidepressants), diazepam (sedatives), amphetamine, (stimulates), and gabapentin (anticonvulsants/mood stabilizers). There was no singular most-prescribed antipsychotic.
Knee pathologies identified in the patients included injuries to the medial meniscus (361 cases), lateral meniscus (371 cases), anterior cruciate ligament (ACL) (313 cases), posterior cruciate ligament (33 cases), medial patellofemoral ligament (34 cases), medial collateral ligament (394 cases), and lateral collateral ligament (470 cases). Holding sex, age, and laterality constant, we observed that patients taking PTMs had lower adjusted odds of presenting with an ACL injury compared with patients not taking PTMs (adjusted odds ratio [aOR], 0.55; 95% CI, 0.32-0.82; P = .003). PTM usage was not found to be significantly associated with any other knee pathology. Antidepressant usage was less commonly seen in patients with ACL injury than in those without ACL injury (aOR, 0.51; 95% CI, 0.27-0.95; P = .033). Similarly, sedative usage was more commonly observed in patients with ACL injury than in those without ACL injury (aOR, 0.44; 95% CI, 0.20-0.98; P = .043). Other drug subclass associations were insignificant at the .05 level.
We found that 245 patients (24.5%) undergoing knee arthroscopy were taking at least 1 PTM in the 6 months prior to surgery. Moreover, we found evidence to suggest that patients taking PTMs are less likely to present with ACL injury compared with patients who are not taking PTMs.
Both increased age and female sex were significantly associated with an increased risk of PTM usage in the 6 months prior to surgery. However, it should be noted that the effect of age was small, given that a 10-year increase in age from the mean age of 34.1 years was only associated with an increase in the probability of PTM usage of 5.9% when all other variables were held constant at baseline (P < .001). Conversely, the effect of patient sex was large, given that female patients used PTMs at higher rates than male patients (odds ratio, 1.92; P < .001). Additionally, female patients were more commonly prescribed multiple PTMs across different classes when compared with male patients. Considering the patient population taking PTMs, 7.0% of female patients used multiple PTMs across different classes versus 3.4% of male patients. Antidepressants were the most common drug class and selective serotonin reuptake inhibitors (e.g., sertraline and fluoxetine) were the most common antidepressant subclass for arthroscopic knee surgery patients.
Contrary to our hypothesis, this study found that PTM use among patients undergoing knee arthroscopy is noticeably lower than that in shoulder and hip arthroscopy populations. Of the 1,000 patients undergoing knee arthroscopy in this study, 245 (24.5%) were taking at least 1 PTM in the 6 months prior to surgery. This rate exceeds the 2019 national average (16.5%) by 8 percentage points.^4^ However, it remains notably lower than previously reported rates in a shoulder arthroscopy population (41.8% reported by Cesarek et al.,^16^ N = 1,279) and hip arthroscopy population (48% reported by Youlo et al.,^8^ N = 880). This finding suggests that the knee arthroscopy population may differ from the hip and shoulder arthroscopy populations across a series of demographic variables associated with PTM usage, such as age and sex.^8^^,^^16^ Consistent with findings by Cesarek et al. and Youlo et al., our data showed higher PTM usage among female patients compared with male patients. In our cohort, 34.2% of female patients used PTMs prior to surgery compared with 19.5% of male patients. This sex-specific disparity is larger than that reported in shoulder patients (55.9% vs 34.1%) and hip patients (53% vs 38%).^8^^,^^16^ Such sex-based differences in PTM usage and the prevalence of psychiatric conditions align with patterns identified in the U.S. population through national epidemiologic studies.17, 18, 19
The reasons patients undergoing knee arthroscopy use PTMs at greater rates compared with the general population have yet to be identified.^8^^,^^16^ Plausible factors include increased risk of joint pathology among patients using PTMs, increased risk for repeated surgery, and demographic differences (i.e., age, sex, and activity level) between the knee arthroscopy population and the general population. Secondary analysis of our study data shows no evidence of either PTM usage linked to an increased risk of repeated knee arthroscopy or PTM usage associated with higher rates of knee injury; contrarily, we found that PTM usage was associated with lower odds of presenting with an ACL injury. However, further research on this topic is warranted.
We did not identify significant year-on-year time trends for either PTM usage overall or drug class usage. High variability in the monthly knee case volume, combined with the low case volume overall, prevented month-on-month time-series analysis from providing viable results. From 2017 to 2021, the percentage of patients taking at least 1 PTM increased from 22.5% to 25.5%, although this increase was not statistically significant. Through a quasi-experimental design, Sanborn et al.^6^ reported declines in the number of dispensed PTMs for the general U.S. population in the initial months of the COVID-19 pandemic. As our analysis was conducted using annual data, we were unable to observe such relationships in the arthroscopic knee surgery population.^6^
There are several limitations to this retrospective study. It is possible that PTM usage was underestimated owing to incomplete patient records. Moreover, it is not known whether patients were actually taking their documented medications, further complicating the interpretation of PTM usage. Also of note, this study did not measure the degree to which patients effectively managed their mental health comorbidities, nor did we identify specific psychiatric diagnoses in the electronic medical records. Instead, we relied on PTM usage as a proxy for psychiatric comorbidity. It is not unreasonable to hypothesize that poorly managed psychiatric comorbidities are associated with worse postoperative outcomes. Although this study focused on preoperative PTM usage as a proxy for psychiatric comorbidity, attention to postoperative PTM usage could have also yielded valuable information. Some patients started PTM regimens within months after surgery, suggesting an undiagnosed psychiatric comorbidity during knee arthroscopy. Regarding the logistic model, additional demographic control variables would have improved statistical modeling of PTM usage, but these variables were unavailable in the patient records. Finally, all clinical records in this study were from a metropolitan institution in a mid-sized city. This is of interest as Edelstein et al.^20^ have shown that PTM usage varies between urban and rural populations.
The rate of PTM usage among patients undergoing knee arthroscopy was 24.5% based on 5 years of records from 1 institution. Female patients had a higher frequency of PTM use compared with their male counterparts.
All authors (Z.R., V.I., N.O.G., S.M.S., W.H., M.K.M.) declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.