Which Interventions Are Effective in Treating Sleep Disturbances After THA or TKA? A Systematic Review

Abstract

Background

Poor sleep quality is a common complaint after total joint arthroplasty (TJA), and it is associated with reports of higher pain and worse functional outcomes. Several interventions have been investigated with the intent to reduce the incidence of postoperative sleep disturbance with varying effectiveness. An aggregate of the best available evidence, along with an evaluation of the quality of those studies, is needed to provide valuable perspective to physicians and to direct future research.

Questions/purposes

In this systematic review, we asked: (1) What is the reported efficacy of the most commonly studied medications and nonpharmacologic approaches, and (2) what are their side effects and reported complications?

Methods

This systematic review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A search using a combination of controlled vocabulary and keywords was performed utilizing Medline (Ovid), Embase (Ovid), Cochrane Central, and Web of Science databases from database inception to 2023, with the last search occurring October 24, 2023, to identify studies that evaluated a sleep intervention on the effect of patient-reported sleep quality after THA or TKA. Inclusion criteria were clinical trials, comparative studies, and observational studies on adult patients who underwent primary TKA or THA for osteoarthritis and who completed validated sleep questionnaires to assess sleep quality postoperatively. We excluded studies on patients younger than 18 years, patients with sleep apnea, TKA or THA because of trauma or conditions other than osteoarthritis, revision TJA, studies in languages other than English, and studies from nonindexed journals or preprint servers. Two investigators independently screened 1535 studies for inclusion and exclusion criteria and extracted data from the included studies. Ultimately, 14 studies were included in this systematic review, including 12 randomized controlled trials and 2 prospective comparative studies. A total of 2469 participants were included, with a mean ± SD age of 65 ± 7 years and 38% men in control groups and 65 ± 7 years and 39% men in intervention groups. Sleep quality questionnaires utilized included the Pittsburgh Sleep Quality Index, Self-Rating Scale of Sleep, 100-mm VAS – Sleep, Sleep Disturbance Numeric Rating Scale, Likert scales, and one institutionally designed questionnaire. Quality analysis was performed utilizing the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Randomized Controlled Trials, where higher scores of 13 indicated a more reliable study, and the Newcastle-Ottawa Quality Assessment Scale for Cohort Studies, where higher scores of 9 indicated a more reliable study and scores < 5 represented a high risk of bias. Two of the randomized controlled trials scored a 12 of 13, and the remaining 10 met every criteria of the JBI checklist. Both comparative studies scored 5 of 9 possible points of the Newcastle-Ottawa Scale.

Results

Melatonin and selective cyclooxygenase-2 inhibitor rofecoxib were found to provide a clinically important benefit to sleep quality within the first postoperative week after TJA. However, rofecoxib was withdrawn from the market globally in 2004 over concerns about increased risk of cardiovascular events. Another cyclooxygenase-2 inhibitor, celecoxib, remains available. No other intervention demonstrated a clinical benefit. Side effects of melatonin include dizziness, headache, paresthesia, and nausea, and it is contraindicated in patients with liver failure, autoimmune conditions, or who are receiving warfarin. Long-term adverse effects of rofecoxib include hypertension, edema, and congestive heart failure, and it is contraindicated in patients with renal insufficiency or who are receiving warfarin. Melatonin is considered safe in older patients, but more caution should be taken with rofecoxib.

Conclusion

Owing to limited evidence in support of most of the interventions we studied, none of these interventions can be recommended for routine use after TJA. Melatonin and rofecoxib may provide a benefit to sleep quality in some patients, but physicians need to understand the adverse effects and contraindications before recommending these interventions. Additionally, rofecoxib is no longer commercially available. Future investigation is warranted to evaluate the effectiveness of interventions with minimal side effect profiles for providers to be able to make an informed decision about interventions for sleep improvement after TJA.

Level of Evidence

Level III, therapeutic study.

Introduction

Total joint arthroplasty (TJA) is one of the most common and effective elective procedures, with approximately 1 million procedures performed annually in the United States [9]. However, sleep quality has been demonstrated to be poor during the early postoperative period following THA and TKA [24]. Patients have reported increased time to fall asleep and increased awakenings in the first 4 to 5 weeks postoperatively, with improvement to above baseline approximately 40 weeks after TJA [24]. Poor sleep is associated with reports of more severe pain and worse functional outcomes after TJA of the hip and knee [7]. As a result, there has been increasing interest in finding interventions to improve sleep quality after TJA to improve postoperative pain and outcomes. Current pharmacologic treatments consist of hypnotics such as zolpidem and melatonin, which have been used to improve sleep quality in isolated capacity [12, 15, 33]. Alternative therapies, such as cognitive behavioral therapy and preoperative meditation, have been studied through clinical trials to modify thought patterns believed to negatively impact sleep disturbances [21]. In reducing the incidence of sleep disturbance, postoperative patients may have enhanced recovery and improved functional outcomes [7].

Several interventions have been investigated with the intent to reduce the incidence of postoperative sleep disturbance with varying effectiveness. These interventions are reported to differ in their indications, reported length of efficacy, contraindications, and adverse effects. An aggregate of studies therefore is needed in order to provide valuable perspective to physicians in selecting an appropriate intervention to improve their patients’ sleep quality as well as to direct future research.

We therefore asked: (1) What is the reported efficacy of the most commonly studied medications and nonpharmacologic approaches, and (2) what are their side effects and reported complications?

Materials and Methods

Search Strategy and Information Searches

This systematic review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A search using a combination of controlled vocabulary and keywords was performed utilizing Medline (Ovid), Embase (Ovid), Cochrane Central, and Web of Science from database inception to 2023, with the last search occurring October 24, 2023, to identify studies that evaluated a sleep intervention on the effect of patient-reported sleep quality after THA or TKA (Appendix 1; http://links.lww.com/CORR/B319).

Although several methods to evaluate sleep quality exist, we elected to utilize only patient-reported sleep quality using validated sleep questionnaires. These were the Pittsburgh Sleep Quality Index (PSQI), Self-Rating Scale of Sleep (SRSS), 100-mm VAS – Sleep (VAS-S), Sleep Disturbance Numeric Rating Scale (SD-NRS), and Likert scales [4, 18, 32, 34]. One study utilized an institutionally designed questionnaire about sleep patterns [5]. These surveys are the most cost-effective and accessible methods to evaluate sleep quality when compared with objective measures like polysomnography [16].

Study Selection

We included studies if they evaluated the effect of an intervention on patient-reported sleep quality after TKA or THA. We did not hand-search bibliographies of included studies or attempt any other methods to find additional studies. Inclusion criteria were clinical trials, prospective, comparative studies, and observational studies on adult patients who underwent primary TKA or THA for osteoarthritis and who completed validated sleep questionnaires to assess sleep quality postoperatively. We excluded studies on patients younger than 18 years, patients with sleep apnea, TKA or THA because of trauma or conditions other than osteoarthritis, revision TJA, studies in languages other than English, and studies from nonindexed journals or preprint servers.

Two reviewers (EP, SVB) independently screened each article for inclusion or exclusion using Covidence systematic review software (Veritas Health Innovation; available at www.covidence.org). Any disagreements between reviewers were resolved by reviewer consensus. In our initial search, we identified 1535 studies. Fourteen met all inclusion criteria after screening the abstract and full text (Fig. 1).

Fig. 1.

This figure shows the PRISMA study selection algorithm used for this study.

Data Extraction and Variables

Two investigators (EP, SVB) independently extracted data from the 14 included studies. The following information was recorded from each article in a Google (Alphabet Inc) spreadsheet: study ID (author last name and year published), study title, authors, start and end dates, country where study was conducted, aim of study, inclusion and exclusion criteria, patient population description, number of participants, number of TKAs and THAs performed, intervention(s) implemented, methods, sleep score and scale, and outcomes. The PSQI was scored from 0 to 21, with higher scores indicating poorer sleep. The SRSS and SD-NRS were scored from 0 to 10, with higher scores indicating poorer sleep [31]. The VAS-S was scored from 0 to 100, and the Likert scales were scored from 0 to 3 or 1 to 4, but higher scores represented either better or worse sleep depending on the study. We evaluated claims of benefit against published minimum clinically important differences (MCIDs) where available. The MCID for the PSQI is ≥ 3 points [6], for the SD-NRS it is 2 to 4 points [32], and for the VAS-S it is 10 points [34]. There is no MCID available in the literature for the SRSS or Likert scales.

Data Analysis

Data extracted from the included studies were analyzed using qualitative synthesis. Studies were grouped by intervention and assessed by their outcome scores and whether they provided a clinical benefit. We did not conduct a meta-analysis as there were too many disparate interventions, and study designs were too dissimilar. We were not able analyze by sex or gender as many source studies did not do so. The included studies contained preliminary data and are therefore at risk of publication bias. We did not assess for publication bias, and we consider it likely to be present, given the emerging nature of this research topic. We performed a detailed review and discussion of the risks and contraindications of each intervention to guide interpretation of the findings despite publication bias.

Quality Assessment

The 12 included randomized controlled trials (RCTs) were appraised to assess the extent to which bias had been addressed in the study design, conduct, and analysis using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Randomized Controlled Trials [1]. This tool is comprised of five domains relating to the validity or quality of the study, including selection and allocation, administration of intervention/exposure, assessment, detection, and measurement of the outcome, participant retention, and statistical conclusion validity [1]. Two reviewers (EP, SVB) independently reviewed each of the 12 RCTs to assess adherence to the checklist components and assigned a final score of 13, with a higher score indicating a more reliable study. Two of the RCTs scored a 12 of 13, and the remaining 10 RCTs met every criterion of the JBI checklist (Table 1).

Table 1.

Assessment of bias using the JBI critical appraisal checklist for RCTs

	Buvanendran et al. [2]	Buvanendran et al. [3]	Clarkson et al. [6]	Fan et al. [10]	Haffar et al. [13]	Li et al. [19]	Lunn et al. [23]	Musclow et al. [25]	Orbach-Zinger et al. [26]	Shafiei et al. [28]	Shakya et al. [29]	Shi et al. [31]
Was true randomization used for assignment of participants to treatment groups?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Was allocation to treatment groups concealed?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Were treatment groups similar at baseline?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Were participants blind to treatment assignment?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Were those delivering treatment blind to treatment assignment?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Unclear
Were outcomes assessors blind to treatment assignment?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Were treatment groups treated identically other than the intervention of interest?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	No	Yes
Was follow-up complete, and if not, were differences between groups in terms of their follow-up adequately described and analyzed?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Were participants analyzed in the groups to which they were randomized?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Were outcomes measured in the same way for treatment groups?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Were outcomes measured in a reliable way?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Was appropriate statistical analysis used?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Was the trial design appropriate, and were any deviations from the standard RCT design (individual randomization, parallel groups) accounted for in the conduct and analysis of the trial?	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Score (of 13 total)	13	13	13	13	13	13	13	13	13	13	12	12

The two prospective comparative studies were assessed using the Newcastle-Ottawa Quality Assessment Scale for Cohort Studies. This tool assesses the quality of nonrandomized studies based on study group selection, comparability of groups, and ascertainment of the outcome. Two reviewers (EP, SVB) independently reviewed the two prospective, comparative studies for adherence to the eight questions and assigned a final score of 9. A higher score indicates a more reliable study, and a score of < 5 represents a high risk of bias [22]. Both studies scored 5 of 9 possible points [5, 19].

Study Characteristics

The average age of participants in the studies included was 65 years in control groups and 65 in intervention groups. The average percentage of men was 38% in control groups and 39% in intervention groups. Two studies investigated melatonin and two investigated pregabalin or gabapentin. The remaining interventions were each investigated by one study and are as follows: selective cyclooxygenase-2 (COX-2) inhibitor (rofecoxib), long-acting oral morphine, nitroglycerin patch, preoperative high-dose methylprednisolone, zolpidem, topical cannabidiol (CBD) and/or essential oils, self-guided meditation, chilled irrigation with epinephrine, removing hip precautions, and electroacupuncture.

Primary and Secondary Study Endpoints

Our primary study goal was to aggregate studies investigating interventions to improve sleep after TJA that demonstrated clinically important benefits. We evaluated clinical importance by comparing results to established MCIDs in the literature when available.

Our secondary study goal was to identify the adverse effects and contraindications of each intervention discussed. We did so by identifying listed adverse effects and contraindications in each study as well as conducting independent research on each intervention.

Results

Efficacy of Pharmacologic and Nonpharmacologic Interventions

The interventions of 50 mg of rofecoxib and 1 mg of melatonin provided clinically important benefits on sleep quality on the first and third postoperative nights after TKA and the first, third, and fifth postoperative nights after THA, respectfully [3, 10]. No other intervention demonstrated a clinical benefit (Table 2).

Table 2.

Interventions and results of included studies

Author	Intervention	Number of participants	Number of THAs	Number of TKAs	Results
Buvanendran et al. [2]	Pregabalin	228	0	228	Treatment with pregabalin provided no benefit to sleep quality on the first postoperative night or overall during hospital admission.
Buvanendran et al. [3]	Selective COX-2 inhibitor (rofecoxib)	70	0	70	Treatment with a COX-2 inhibitor provided a clinically important benefit to sleep quality on the first and third postoperative nights.
Canfield et al. [5]	Self-guided meditation	380	0	380	Self-guided meditation provided no benefit to sleep quality after 2 weeks postoperatively.
Clarkson et al. [6]	Melatonin	118	52	66	Patients treated with melatonin reported no difference in sleep quality 2 and 6 weeks postoperatively.
Fan et al. [10]	Melatonin	139	139	0	Treatment with 1 mg of melatonin provided a clinically important benefit to sleep quality on postoperative nights 1, 3, and 5.
Haffar et al. [13]	Topical CBD and/or essential oils	80	0	80	Patients treated with CBD and/or essential oils reported no difference in sleep quality up to 42 days postoperatively.
Li et al. [19]	Chilled irrigation with epinephrine	389	0	389	Chilled irrigation with 0.5% epinephrine provided no benefit to sleep quality.
Lightfoot et al. [20]	Removing hip precautions	237	237	0	Patients with and without hip precautions reported no difference in sleep quality 6 weeks and 3 months after THA.
Lunn et al. [23]	Gabapentin	299	0	299	High- and low-dose gabapentin regimens provided no benefit to sleep quality on postoperative nights 1-6.
Musclow et al. [25]	Long-acting oral morphine in addition to a routine postoperative regimen	190	108	82	Long-acting oral morphine provided no benefit to sleep quality on postoperative nights 1-4.
Orbach-Zinger et al. [26]	Nitroglycerin patch	29	0	29	Patients treated with a nitroglycerin patch reported no differences in sleep quality on the first postoperative night.
Shafiei et al. [28]	Preoperative high-dose methylprednisolone	70	70	0	High-dose methylprednisolone administered intraoperatively provided no benefit to sleep quality after 2 weeks postoperatively.
Shakya et al. [29]	Zolpidem	160	160	0	Zolpidem provided no benefit to sleep quality after 3 months postoperatively.
Shi et al. [31]	Electroacupuncture	80	0	80	Electroacupuncture provided no benefit to sleep quality after 72 hours postoperatively.
Total		2469	766	1703

Melatonin

One study found a large difference between patients who had undergone THA who took 1 mg of melatonin 1 day preoperatively and 5 days postoperatively in VAS-S scores after 1, 3, and 5 days postoperatively [10]. Unfortunately, the data are plotted graphically, and exact scores are not reported. There is clearly a difference of > 10 points between the two groups on postoperative days 1, 3, and 5, which surpasses the MCID for the VAS-S [34]. Another study found no difference in sleep quality between patients who took 6 mg of melatonin and those who took a placebo at 2 and 6 weeks postoperatively after both TKA and THA [6]. These similarly sized studies were both well organized and executed, but they had conflicting results. The former [10] focused on early postoperative sleep quality after THA, while the latter [6] evaluated late postoperative sleep quality after both THA and TKA. More studies are needed to investigate the efficacy of melatonin on postoperative sleep quality after TKA. However, these preliminary results do suggest that melatonin provides a clinical benefit to sleep quality within the first week after THA.

Pregabalin/gabapentin

Two studies found no clinical benefit on sleep quality after TKA to a regimen of 300 mg of pregabalin 1 to 2 hours before surgery, 150 mg twice daily the first 10 postoperative days, 75 mg twice daily on days 11 and 12, and 50 mg twice daily on days 13 and 14, or to two potential regimens of 1300 mg per day of gabapentin or 900 mg per day of gabapentin [2, 23].

Zolpidem

A study investigating the effect of 10 mg of zolpidem administered 2 days preoperatively to 5 days postoperatively 30 minutes prior to bedtime found that patients in the zolpidem group reported better scores in all seven domains of the PSQI 3 weeks after THA and in five domains of the PSQI 3 months after THA compared with the placebo group [29]. However, there is no established MCID for PSQI subcomponent scores, and overall PSQI scores were not reported. Therefore, the clinical benefit of zolpidem on sleep quality after TJA warrants further investigation.

Morphine

One study investigating long-acting oral morphine in addition to a routine postoperative analgesic regimen found no benefit to a regimen of 30 mg of morphine sulfate every 12 hours for 3 days on sleep quality as reported on a 4-point Likert scale for the first 4 postoperative days after TKA and THA [25].

COX-2 Inhibitor

One study investigating the selective COX-2 inhibitor rofecoxib found that patients who received 50 mg of rofecoxib 24 hours preoperatively, 50 mg for 5 days postoperatively, and 25 mg for 8 more days had better median [IQR] SD-NRS scores compared with the placebo group on the first (0.5 [0 to 4.5] versus 5 [1.5 to 8.5]; p = 0.006) and third (0 [0 to 0] versus 4 [2 to 5]; p < 0.001) postoperative nights after TKA [3]. These differences surpassed the MCID of 2 to 4 points, indicating that rofecoxib provides a clinically important benefit in sleep quality after TKA.

CBD

One small study found no benefit to utilizing topical CBD, topical essential oils, or combination CBD and essential oils at 0, 1, 2, 7, 14, and 42 days after TKA, as assessed by the VAS-S score [13].

Nitroglycerin

One study analyzing the effects of transdermal nitroglycerin as an adjuvant to morphine after TKA found no difference in sleep quality 24 hours postoperatively compared with the placebo group [26].

Methylprednisolone

One study evaluating the effects of high-dose (125 mg) intravenous methylprednisolone given to patients preoperatively found that a larger proportion of patients in the treatment group reported PSQI scores of ≤ 5 (74.3% [26 of 35]) compared with the placebo group (31.4% [11 of 35]; p = 0.001) 2 weeks after THA [28]. Although the differences in reported good sleep quality between the treatment and placebo groups are large, overall PSQI scores were not provided, so we cannot assess the clinical importance of the differences in sleep quality.

Nonpharmacologic Interventions

One study investigating hip precautions found no difference in sleep quality between patients utilizing standard hip precautions compared with those not utilizing any precautions at 6 weeks and 3 months after THA [20].

Meditation, implemented with a self-guided meditation video twice a day beginning 2 weeks prior to TKA and ending 2 weeks postoperatively, found differences in bedtime, hours slept, and nighttime awakenings utilizing an institutionally designed questionnaire between those who meditated and those who did not [5]. However, as there is no MCID established for this questionnaire, we cannot assess the clinical importance of these differences.

One study evaluating the use of intraoperative cold irrigation using 4000 mL of 4°C cold saline with 0.5% of epinephrine during TKA with the intention of prolonging the cryotherapy process found no clinically important benefit to sleep quality [19].

Electroacupuncture (acupuncture with an electric current through the needles) was associated with better mean ± SD SRSS scores compared with controls at 24 (3.5 ± 0.6 versus 4.3 ± 0.7; p < 0.05), 48 (2.8 ± 0.7 versus 3.7 ± 0.6; p < 0.05), and 72 (1.7 ± 0.7 versus 3 ± 0.7; p < 0.05) hours after TKA [31]. However, these effect sizes are small, and as there is no available MCID for the SRSS, we are unable to determine the clinical benefit of electroacupuncture on sleep quality after TKA.

Side Effects and Reported Complications

Melatonin

Side effects of melatonin include dizziness, headache, paresthesia, and nausea. In the study by Fan et al. [10], there was no difference in side effects between the treatment and control groups. Neither study investigating melatonin reported any complications [6, 10].

Pregabalin/gabapentin

Adverse effects of pregabalin and gabapentin include sedation, confusion, dizziness, headache, dry mouth, peripheral edema, and diplopia [2, 23]. In the study evaluating pregabalin, sedation, confusion, and dry mouth occurred more frequently in the treatment group than the placebo group on the day of surgery and first postoperative day, but not thereafter. If sedation occurred, they compensated by reducing basal epidural flow rate [2]. There were no falls or other complications [2]. In the study evaluating gabapentin, dizziness was more pronounced in the high-dose (1300 mg) and low-dose (900 mg) gabapentin groups from postoperative days 1 to 6. The high-dose (1300 mg) gabapentin group had a higher incidence of visual disturbances in the first 2 days postoperatively compared with the placebo group [23]. Lunn et al. [23] cautioned against the standard use of gabapentin in older patients as sedation and dizziness may impede early mobilization and cause falls.

Zolpidem

The study investigating zolpidem did not report on contraindications, side effects, or adverse effects [29]. However, the drug has been associated with several side effects, including hallucinations and sensory distortions, amnesia, sleepwalking/somnambulism, and nocturnal eating. Additionally, caution should be taken when prescribing zolpidem to female and older patients. The former have been found to have higher serum zolpidem concentrations than male patients after receiving the same dose, and the latter have lower clearance and volumes of distribution of the drug. Contraindications include patients with low albumin, as zolpidem is highly protein bound, and concomitant administration with cytochrome P450 inhibitors, which may increase zolpidem serum levels [14].

Morphine

Opioids are associated with several side effects, including nausea, vomiting, pruritis, drowsiness, dizziness, and constipation [25]. Additionally, there is the risk of oversedation and overdose. In the study evaluating the addition of long-acting oral morphine to a routine regimen, only nausea and vomiting were increased in the treatment group on postoperative day 3 compared with the control group. However, 10 patients in the treatment group had to stop all opioids for confusion and/or oversedation compared with three in the usual care group (p = 0.08) [25].

COX-2 Inhibitor

Rofecoxib was withdrawn from the market globally in 2004 over concerns about increased risk of cardiovascular events. Another cyclooxygenase-2 inhibitor, celecoxib, remains available. While COX-2 inhibitors minimize the gastrointestinal and bleeding adverse effects seen with NSAIDs, they still carry the same risk of renal toxicity and have been associated with the long-term adverse effects of hypertension, edema, and congestive heart failure [30]. There is also a possibility of a small potentiation of warfarin effect. The study evaluating rofecoxib found no differences in international normalized ratio or prothrombin time between treatment groups, and no patients had any bleeding complications [3]. No other adverse effects or complications were reported in this study, but caution should be taken when considering this medication for patients with renal insufficiency and known coagulation abnormality [3].

CBD

Potential side effects of topical CBD include skin reactions. There were no differences in adverse effects of complications between treatment and control groups in the study evaluating topical CBD and essential oils [13].

Nitroglycerin

In previous studies, no serious side effects of nitroglycerin patches have occurred. However, a few studies have shown that nitroglycerin may cause rebound myocardial ischemia after patch removal. Indeed, two patients in the included study evaluating this intervention suffered from myocardial infarctions 12 hours after the patch was removed, prompting institutional review board shutdown of the study [26]. Extreme caution should be taken when considering this pharmacologic intervention, especially in patients with a history of cardiac disease.

Methylprednisolone

Corticosteroid use is associated with risks, including hyperglycemia (especially in patients with diabetes), increased blood pressure, wound dehiscence, deep vein thrombosis, and infection [28]. A single administration of high-dose methylprednisolone in Shafiei et al. [28] was associated with higher systolic blood pressure in the treatment group on postoperative day 1, but none of the patients required additional blood pressure medication. No other adverse effects were reported; however, patients with comorbidities such as diabetes were excluded, so further investigation is warranted in more diverse patient groups [28].

Nonpharmacologic Interventions

The nonpharmacologic interventions have fewer potential side effects and complications. Hip precautions, designed to prevent postoperative dislocation, have been associated with slowed recovery and return to activity, additional expenses, and decreased patient satisfaction [20]. The study investigating the use of hip precautions found no difference in complications, including dislocation, between the group that adhered to the precautions compared with those that did not [20]. Meditation is not associated with any known side effects, and the study evaluating meditation found no difference in adverse events or complications between the groups that utilized meditation compared with those that did not [5]. Cold irrigation was not associated with any side effects or complications either [18]. The study evaluating electroacupuncture reported no side effects or complications and did not stipulate any contraindications for the treatment [31].

Discussion

Sleep quality has been demonstrated to be poor in the postoperative period after TJA, leading to increased pain and worse functional outcomes [7]. Several interventions have been investigated with the intent to reduce the incidence of postoperative sleep disturbance with varying effectiveness, contraindications, and side effects. An aggregate analysis of studies is needed to provide valuable insight to physicians in selecting an appropriate intervention to improve patients’ sleep quality and to direct future research. We identified two interventions that provided a clinically important benefit to patients’ sleep quality: 50 mg of selective COX-2 inhibitors after TKA and 1 mg of melatonin after THA. Sleep quality was only evaluated within the first postoperative week in both studies. Further research is needed to evaluate these interventions for a more prolonged postoperative period and for effectiveness in both TKA and THA.

Limitations

This systematic review has limitations, many of which are inherent to the studies examined. First, we were only able to identify one to two studies on each intervention due to the recent interest in this topic. For that reason, our evidence base consists primarily of preliminary studies. Such preliminary evidence often is prone to publication bias (positive outcome bias), meaning it may overstate benefits and underreport harms. To account for this, we conducted a comprehensive investigation into the contraindications, side effects, and complications of each included intervention.

Second, we did not conduct an analysis by sex or gender, and our findings cannot necessarily be applied equally to male and female patients (or men and women). To circumnavigate this limitation, we have discussed differences in intervention effectiveness by sex and gender when possible and relevant in the contraindications and adverse effects section of the results.

Third, many of the studies have short follow-up times. Only six of the studies reported sleep quality ≥ 2 weeks postoperatively. However, poor sleep quality for even just the first few nights after surgery is associated with longer hospital stays, increased pain sensitivity, and altered cognitive function [27].

Fourth, we performed a qualitative rather than a quantitative data analysis. Given the varied interventions studied and different questionnaires utilized to assess outcomes, the data were too heterogeneous for a quantitative analysis. Our study provides a preliminary look into an emerging evidence base, and we therefore invoke caution in the interpretation of these findings. The benefit to this review of heterogeneous studies was that we were able to critically appraise the quality of evidence and clinical importance of findings. Most of the studies found statistically significant correlations that did not reach clinical significance. Readers can use our systematic review to evaluate the interventions with a more critical view of the evidence, benefits, and harms than was often provided in the initial study.

Fifth, no study directly correlated sleep quality with pain scores or any other potential causes of sleep disturbance. Sleep disturbances after surgery have many causes, and sleep can be influenced by pharmacology, perceived pain, and environmental factors [27]. With studies performed in a variety of settings, it is more challenging to directly correlate an intervention’s effectiveness in reducing sleep disturbances. However, every included study utilized a control group to compare the intervention group considering the multifactorial causes of sleep dysfunction. Given the movement toward shorter lengths of hospital stay for surgical procedures, it is important to consider changes if differences in sleep environment (such as at home versus in the hospital) factor into overall sleep quality. However, variability in hospital procedures as well as various methods used by studies to record sleep disturbances both at home and in the hospital limit the extent to which this information can be generalized and require further study and evaluation.

Finally, we evaluated subjective, patient-reported outcomes rather than objective measures such as polysomnography. Polysomnography is the most accurate assessment of sleep quality, quantity, and architecture, but as it requires an overnight stay in a sleep clinic or laboratory, it makes long-term recordings impractical [17].

Discussion of Key Findings

We found only one study on most of the treatments we evaluated, and almost none of the “benefits” claimed by the source studies exceeded the MCID where one is defined, or even a modest effect size (where there is no defined MCID). Since all drugs carry potential risks, none of these interventions can be recommended for routine use. The two interventions that did find a clinically important benefit were rofecoxib and melatonin. Melatonin’s side-effect profile is relatively minimal, with risk of dizziness, headache, paresthesia, and nausea [10]. Melatonin’s relative contraindications include liver failure, autoimmune conditions such as rheumatoid arthritis, and concurrent use with warfarin. Melatonin is considered safe for older patients [8]. COX-2 inhibitors, such as rofecoxib, minimize gastrointestinal and bleeding adverse effects associated with NSAID use. However, they do carry the same risk of renal toxicity as NSAIDs and are therefore contraindicated in patients with renal insufficiency [3]. Rofecoxib, like melatonin, is also contraindicated in patients receiving warfarin due to a small risk of potentiation of warfarin’s effects [3]. In the long term, they carry risk of hypertension, edema, and congestive heart failure. Notably, rofecoxib was withdrawn from the market globally in 2004 over concerns about increased risk of cardiovascular events. Celecoxib is another COX-2 inhibitor with a similar mechanism of action but with lower cardiovascular risk, and is the only COX-2 inhibitor available in the United States. The effect of celecoxib on sleep quality has yet to be studied. Before prescribing or recommending these pharmaceuticals, providers need to consider their patient’s medical histories, relative and absolute contraindications, and engage in discussion about potential adverse effects with patients.

None of the nonpharmacologic interventions demonstrated a clinical benefit on postoperative sleep quality. While there were differences in reported sleep quality in patients who adhered to a self-guided meditation regimen compared with those who did not, these differences did not exceed the MCID. The evidence of self-guided meditation in the included study was weak, with nonrandomized treatment and control groups and outcome assessment utilizing an institutionally designed questionnaire. Some studies have demonstrated the efficacy of meditation in combating insomnia, with effectiveness even postoperatively [11, 21]. As such, this intervention warrants further evaluation with a higher quality randomized controlled study. Intraoperative cold irrigation with epinephrine also resulted in differences in postoperative sleep quality that did not exceed the MCID. This was a large, well-conducted study, that only tracked sleep quality for one postoperative night after TKA. This intervention merits further investigation with a longer follow-up period.

A cautious approach should be taken when prescribing medications postoperatively. Although good sleep quality has been shown to positively impact the outcomes of TJA, it is important to weigh the risks, benefits, and contraindication of each intervention. The risks often outweigh the benefits in several of the described pharmaceuticals, most of which did not even demonstrate a clinical benefit. Methylprednisolone and zolpidem, for example, are both contraindicated in older patients due to decreased adrenal function for the former and cognitive dysfunction and respiratory depression for the latter. As the population of patients receiving TJA is generally of this age range, it would be harmful to recommend using these drugs as routine sleep aids. On the other hand, interventions with a minimal side-effect profile, such as melatonin and rofecoxib, merit further investigation in patients who have undergone TKA or THA with longer follow-up. Further research would enable providers to make an informed decision about which interventions they can use to safely improve patient sleep quality so patients may achieve the best possible outcomes after TJA.

Conclusion

Sleep has been well established as a variable that impacts pain and outcomes after TJA. Most reported interventions provided no clinical benefit to sleep quality after TJA. Only a selective COX-2 inhibitor and melatonin were demonstrated to have a clinically important benefit within 1 week after TKA and THA, respectively. These interventions require further investigation for effectiveness after both TKA and THA and for a longer postoperative period. Selective COX-2 inhibitors and melatonin may provide benefits to sleep quality in some patients, although they should be prescribed only with caution in patients receiving warfarin or with liver failure, renal insufficiency, heart failure, hypertension, and autoimmune conditions. Additionally, rofecoxib is no longer commercially available. The other described interventions provided no clinical benefit to patients, and because of the potential adverse effects, especially among older patients, these interventions are not recommended for routine use to improve sleep quality after TJA. Further investigation is needed to provide more options for safely improving sleep quality after TJA.

Supplementary Material

abjs-483-105-s001.docx ^{(24.2KB, docx)}