Secondary Logo

Journal Logo

Economics, Education, and Policy: Research Reports

Performance Characteristics and Validation of the Opioid-Related Symptom Distress Scale for Evaluation of Analgesic Side Effects After Orthopedic Surgery

YaDeau, Jacques T. MD, PhD; Liu, Spencer S. MD; Rade, Matthew C. BA; Marcello, Dorothy BA; Liguori, Gregory A. MD

Author Information
doi: 10.1213/ANE.0b013e31821ae3f7

The side effects of opioids may limit their use and acceptability for postoperative analgesia. Peripheral nerve blockade, epidural analgesia, and analgesic adjuncts are often used both to alleviate pain and to diminish the incidence and severity of opioid-related symptoms. Reduction in opioid use can be used as an end point for analgesic trials, but opioid dose is a surrogate measure that is not necessarily linearly related to occurrence of opioid-related symptoms.1 Variation in opioid type, dose, and route, as well as administration of analgesic adjuncts, antiemetics, and local anesthetic, may influence opioid-related symptomatology in a complex manner. Examination of multiple symptoms potentially attributable to opioids may provide useful information for clinical trials; instruments that measure only one symptom underestimate the impact of other symptoms.2

The Opioid-Related Symptom Distress Scale (ORSDS) uses 4-point Likert scales to evaluate the frequency, severity, and bothersomeness of 12 symptoms: nausea, vomiting, constipation, difficulty passing urine, difficulty concentrating, drowsiness or difficulty staying awake, feeling lightheaded or dizzy, feeling confused, feelings of general fatigue or weakness, itchiness, dry mouth, and headache (see Appendix).3 The symptom-specific ORSDS score is the average of the 3 symptom distress dimensions (frequency, severity, and bothersomeness). The composite ORSDS is the mean of the 12 individual symptom-specific scores. Potential ORSDS scores range from 0 to 4. The ORSDS was adapted from the Memorial Symptom Assessment Scale, which provides quantitative data about a large group of symptoms to monitor clinical changes after therapeutic interventions.4 The Memorial Symptom Assessment Scale assesses 32 symptoms using 4- to 5-point Likert scales and dimensions of frequency, severity, and distress or bothersomeness. Dimensionality analysis concluded that combining distress with either frequency or severity (depending on the symptom) yielded significantly more information than considering distress alone.

The ORSDS has been validated in postoperative day (POD) 1 outpatient laparoscopic cholecystectomy performed under general anesthesia (GA) in patients receiving fentanyl and cyclooxygenase-2 inhibitors.3 Additional validation studies were suggested in other patient populations. To our knowledge, the ORSDS has not been validated for orthopedic patients. In this study, we investigated performance characteristics and validity of the ORSDS on POD 1 in patients receiving 4 different anesthetic and analgesic regimens after orthopedic surgery. Inpatients were also evaluated on POD 3. Some patients in the study received GA and others regional anesthesia. Postoperative analgesia ranged from oral analgesics to peripheral nerve blockade (patient-controlled epidural analgesia [PCEA]), epidural, or IV patient-controlled analgesia (PCA).

This validation study used a patient sample receiving routine care at the authors' institution and was not primarily intended to compare clinical data with other studies. Validation is a valuable prerequisite for use of the ORSDS in clinical trials assessing efficacy and side effects of analgesics. An instrument that is valid for one patient population may not be valid for a different patient population. Demonstration of validity of the ORSDS among patients undergoing a wide range of anesthetic and analgesic regimens would be useful because validation offers assurance that for these patients the ORSDS actually measures what it purports to measure. Additionally, systematic reviews and meta-analyses of analgesic therapies can only compare data among studies when the studies report the same data. We speculate that widespread use of standard, validated outcome measures for opioid-related side effects would facilitate systematic comparison of analgesic trials. It was hypothesized that the ORSDS would be validated in these 4 groups of patients undergoing diverse anesthetic and analgesic regimens. This validation study was not intended to demonstrate the utility of the ORSDS, because that would require further study.

Validation of an instrument for scientific study can be straightforward if there is an available objective measure or a widely accepted alternative instrument. Unfortunately, there are no objective measures or widely accepted alternative instruments to measure opioid-related symptom distress. In the absence of an objective “gold standard,”5 validation of an instrument is the degree to which the evidence supports the intended interpretation of the instrument.6 Similar to its progenitor, the Memorial Symptom Assessment Scale, the ORSDS is intended to provide quantitative data about a set of symptoms to monitor clinical changes after therapeutic interventions. Validation of the ORSDS was performed by assessing 7 characteristics: internal consistency, content validity, construct validity, principal components analysis, known group validity, responsiveness, and opioid dose dependency.3 To determine eligibility for the study, the type of surgery was used as a proxy for the anesthetic and analgesic regimen provided. Depending on the operative site, the anesthetics consisted of peripheral nerve blockade, neuraxial anesthesia or GA, and either oral analgesics alone or oral analgesics in combination with IV opioids, analgesic nerve blockade, epidural analgesia, or IV opioids. This study was not intended primarily to compare efficacy or side effects among these different regimens, but rather to validate the ORSDS as a tool for future studies that could address these important problems.

METHODS

This study was approved by the IRB of Hospital for Special Surgery. The ORSDS (as well as validation questions) was administered to 4 groups of orthopedic patients after obtaining informed written consent. Peripheral nerve blockade (Peripheral) of the brachial plexus, performed for upper extremity surgery (surgical site at or below the elbow), was a model for ambulatory orthopedic surgery with low levels of postoperative pain expected. Neuraxial anesthesia (Neuraxial), performed for anterior cruciate ligament reconstruction with allograft, was a model for ambulatory orthopedic surgery with moderate pain. Combined spinal-epidural anesthesia/femoral nerve block (FNB)7/PCEA (Regional), performed for unilateral primary total knee arthroplasty, was a model for inpatient orthopedic surgery with moderate-to-severe pain. PCEA (bupivacaine 0.06% + hydromorphone 10 μg/mL) was continued until noon of POD 2. GA/IV hydromorphone PCA (GA), performed for 1 to 2 level posterior lumbar spine fusion, was a model of inpatient orthopedic surgery with moderate-to-severe pain.

Oral and IV opioids were given in the postanesthesia care unit as needed. Ambulatory patients were prescribed oral opioids for postoperative analgesia. For both inpatient groups, oral opioids were usually initiated before discontinuation of the PCA. Exclusion criteria included failure to follow the anesthetic plan, patients younger than 18 years, dementia, and communication issues such as severe hearing impairment or diminished ability to communicate in English (translations would have to be separately validated). Patients were also excluded if they received postoperative ketamine infusions. The study was completed with 50 patients in each group (200 total). Because this study was meant as a validation and feasibility study, sample size and power analyses were not relevant.

Data Collection

Ambulatory patients were contacted by telephone on POD 1; hospitalized patients were interviewed in person on PODs 1 and 3. Initial administration of the survey to patients on POD 1 was attempted 24 hours after surgery, but this was not always feasible. Some inpatients had conflicting therapeutic measures (e.g., physical therapy), and some outpatients preferred to be interviewed at other times. Survey administration on POD 3 generally occurred between 9:30 and 11 AM, before potential discharge.

The ORSDS evaluated each of the 12 symptoms separately (Appendix) by asking the patient about the dimensions of frequency, severity, and bothersomeness. Each dimension was given a score from 0 to 4, in which 0 represented “did not experience” the symptom. Frequency, if present, was graded from “rarely” to “almost constantly.” Severity was graded from “slightly” to “very.” Bothersomeness was graded from “not at all” to “very much.” The “symptom-specific ORSDS” was the mean of patient-reported scores for the 3 dimensions of distress. The “composite ORSDS” was the mean of all 12 symptom-specific scores. Potential ORSDS scores ranged from 0 to 4.

Analgesic dosages were obtained via hospital records and by interview. Epidural opioid use was converted to oral morphine equivalents at a rate of 1 mg epidural hydromorphone to 30 mg oral morphine. This includes epidural hydromorphone converted to IV hydromorphone at a ratio of 1:2,8 IV hydromorphone converted to IV morphine at a ratio of 2:10,9 and IV morphine converted to oral morphine at a ratio of 1:3.9 Subcutaneous hydromorphone converted to oral morphine 1:15.9 Oral morphine conversion ratios were as follows: oxycodone 1 mg:1.5 mg oral morphine,10 propoxyphene 200 mg:30 mg,11 hydrocodone 1:1,10 and hydromorphone 1:4.9

Data Analysis

Performance analysis evaluated usefulness of the ORSDS as an outcome for trials after orthopedic surgery. Descriptive statistics were used. Validation was modeled on the initial validation of the ORSDS. It was hypothesized that the ORSDS would have a similar validation profile as was found for outpatient cholecystectomy. P values were unadjusted for multiple testing.

Validation-related questions derived from a Health Outcomes Recovery Questionnaire included questions about clinically meaningful events (CMEs).3 CMEs for each dimension were defined in the following manner: side effects reported by the patient as occurring frequently to almost constantly, side effects reported as moderate to very severe in strength, and side effects reported by the patient as bothering them quite a bit to very much. Global measures included general satisfaction (7-point scale), maximum activity level (5-point scale), percent (%) normal activity, and hours of assistance needed. Symptom-specific measures were used to validate relevant ORSDS scores. Numeric Rating Scale (NRS) pain scores were obtained by asking patients their pain level on a 0 to 10 scale.

Multivariable linear regression analyses (performed with the SPSS data editor, version 14.0 for Windows; SPSS, Inc., Chicago, IL) determined whether opioid intake correlated with either symptom-specific or composite ORSDS scores. For analyses we used (1) the ORSDS data and opioid intake for POD 1, (2) POD 3 data (for Regional and GA), and (3) the composite ORSDS scores and cumulative opioid intake, POD 0 to POD 3. Patient age, body mass index, race (non-Caucasian), and gender (female) were also used as independent variables in each initial regression model with a stepwise methodology. Independent variables were excluded from the model if their significance factor was >0.075. The regressions were then repeated using each symptom-specific ORSDS score (e.g., nausea, vomiting, constipation) as the dependent variable, respectively. One-way analysis of variance was performed for NRS pain scores and opioid use to determine whether there were any differences among the 4 procedure groups.

RESULTS

Patient Characteristics

To have 50 patients per group, the following numbers were enrolled: 70 Peripheral, 66 Neuraxial, 52 Regional, and 63 GA (Table 1). Mean patient age ranged from 50 years (Peripheral) to 64 years (Regional), and the mode for ASA physical status ranged from I to II (Table 2).

Table 1
Table 1:
Patient Enrollment and Exclusions
Table 2
Table 2:
Demographic Characteristics and Clinical Features

Performance Analysis

Opioid use varied significantly among the 4 patient groups (Table 3, P < 0.0001). Patients used opioids equivalent to the following milligrams of oral morphine through POD 1: Peripheral, 28 mg (15:53) (median [25 percentile:75 percentile]); Neuraxial, 60 mg (38:75); Regional, 71 mg (48:105), 51% from oral or subcutaneous administration; and GA, 190 mg (111:282), 24% from oral or subcutaneous administration. Opioid consumption through POD 3 was as follows: Regional, 171 mg (125:238), 74% from oral or subcutaneous administration; and GA, 355 mg (199:529), 49% from oral or subcutaneous administration.

Table 3
Table 3:
Opioid Usage and Pain Scores

Average NRS pain scores ranged from moderate to high: 3.9 ± 2.9 (Peripheral); 4.2 ± 2.0 (Neuraxial); 3.3 ± 2.4, 4.1 ± 2.1 (Regional; POD 1 and POD 3, respectively); and 4.7 ± 2.5 and 4.3 ± 2.1 (GA; POD 1 and POD 3, respectively). Variance of pain scores among groups was not significant (P = 0.09), but the study may have been underpowered for this comparison. A comparison of NRS pain scores for Regional on POD 1 and POD 3 gave a P value of 0.007 (paired samples t test). We speculate that the increased pain on POD 3 may reflect discontinuation of PCEA (on POD 2), diminishing effects of the FNB (which improves pain scores for these patients on POD 1 and POD 2),7 and increased movement associated with progression of physical therapy. The number of CMEs increased parallel to opioid consumption (Table 4).

Table 4
Table 4:
Number of Clinically Meaningful Events Reported

There is not yet an accepted system for analysis of ORSDS scores. The median composite ORSDS score allows a simple comparison between groups. For many data points, the symptom-specific score is 0 (indicating “did not experience”); the data tend to be skewed and the mean ORSDS score may be excessively influenced by outliers. Examination of symptom-specific ORSDS scores allows the researcher or clinician to determine which symptoms cause distress. One could focus on symptoms with a median symptom-specific ORSDS score >0. However, in some cases (e.g., Peripheral), there are few or no symptoms for which at least half of the subjects report distress. This analysis ignores the possibility that a substantial minority of patients may be experiencing opioid-related symptom distress. We speculate that problematic symptoms can be identified by determining which symptoms have upper quartile symptom-specific ORSDS scores >1. This means that at least 25% of patients rate this symptom with an average score >1. Symptom-specific dimensional scores >1 are described with a frequency of at least “occasionally,” a severity of at least “moderately,” and bothersomeness level of at least “a little bit” (Appendix).

Median and quartile ORSDS scores are displayed (Figs. 1 and 2) for the composite ORSDS scores and the 6 most frequently reported symptom-specific ORSDS scores (nausea, drowsiness, dizziness, fatigue, itchiness, and dry mouth). Additional ORSDS data can be found in the Supplemental Digital Content 1, http://links.lww.com/AA/A276. For Peripheral patients, the median composite ORSDS score was 0.19, and no symptom had a median score >0. Symptoms with a 75 percentile (upper quartile) symptom-specific ORSDS score >1 (i.e., at least 25% of patients were distressed by the symptom) were drowsiness, dry mouth, and fatigue.

Figure 1
Figure 1:
Median Opioid-Related Symptom Distress Scale scores. POD = postoperative day; GA = general anesthesia.
Figure 2
Figure 2:
Median Opioid-Related Symptom Distress Scale Scores. POD = postoperative day; GA = general anesthesia.

Neuraxial patients displayed a larger array of distressing symptoms. The median composite ORSDS score was 0.52, and a median symptom-specific score >0 occurred for drowsiness and dry mouth. Symptoms with a 75 percentile symptom-specific ORSDS score >1 were nausea, drowsiness, feeling dizzy, itchiness, dry mouth, and fatigue.

On POD 1 for Regional, the median composite ORSDS score was 0.51; symptoms with median symptom-specific ORSDS scores >0 were dizziness and dry mouth. At least 25% of Regional patients on POD 1 had symptom-specific ORSDS scores >1 for nausea, drowsiness, dizziness, itchiness, dry mouth, and fatigue. On POD 3, dry mouth and fatigue had median symptom-specific ORSDS scores >0. The list of POD 3 distressing symptoms (75 percentile ORSDS scores >1) was similar to POD 1, except that nausea had been replaced by constipation, and difficulty concentrating emerged as a problem.

For GA on POD 1, the median composite ORSDS score was 0.94. Symptoms with median symptom-specific ORSDS scores >0 were nausea, drowsiness, dizziness, dry mouth, and fatigue. Symptoms with a 75 percentile symptom-specific ORSDS score >1 on POD 1 were nausea, difficulty concentrating, drowsiness, dizziness, itchiness, dry mouth, fatigue, and headache. On POD 3, drowsiness, dry mouth, and fatigue had median symptom-specific ORSDS scores >0. On POD 3, constipation was added to the list of distressing symptoms (75 percentile symptom-specific ORSDS scores >1).

Validation Metrics

Validation was performed in the same manner as the initial validations of the ORSDS3: by assessing internal consistency, content validity, construct validity, component analysis, known group validity, responsiveness, and correlation of scores with opioid intake. A summary of the validation follows (for detailed results please see Supplemental Digital Content 1 [http://links.lww.com/AA/A276 and Tables A–U [see Supplemental Digital Content 2, http://links.lww.com/AA/A277).

Internal consistency was evaluated using Cronbach's coefficient α.6 Internal consistency measures the degree of interrelationship between a set of items. Low alphas may indicate a lack of homogeneity, but excessively high alphas may be associated with a narrow scale with diminished validity. Internal consistency (Cronbach coefficient α) was high for each patient group, ranging from 0.75 to 0.99, comparing the 3 dimensions, for symptom-specific and composite ORSDS scores.

Content validity refers to the extent of coverage of relevant concepts; it reflects the extent to which a test includes all the items needed to represent the measured concept. Content validity for the ORSDS was addressed by the previous validation.3 Content validity was also supported by the finding that high symptom-specific ORSDS scores were associated with adverse CMEs (P < 0.001) (Tables A–D, see Supplemental Digital Content 2, http://links.lww.com/AA/A277).

Construct validity is not a static measurement, but rather a testing of hypothesized relationships between aspects of the scale and observable criteria specified by a theoretical system. Construct validity (whether scores correlate with measures of other variables in the hypothesized way) was evaluated by comparing ORSDS scores with general and symptom-specific validation measures. It was supported by the finding that the ORSDS scores (composite and some symptom-specific) were related to activity levels for Peripheral (composite P = 0.0003), Regional (composite P = 0.03), and GA (symptom-specific P = 0.01) (Tables E–H, see Supplemental Digital Content 2, http://links.lww.com/AA/A277). We did not find that hours of assistance needed corresponded to composite ORSDS scores (Table I–L, see Supplemental Digital Content 2, http://links.lww.com/AA/A277). Patient satisfaction was significantly related to some symptom-specific ORSDS scores, with the related factors varying by patient subgroup. Satisfaction was significantly related to composite ORSDS scores for Regional on POD 3 and GA on POD 1 (Tables M–P, see Supplemental Digital Content 2, http://links.lww.com/AA/A277).

Exploratory principal components analysis was performed with symptom-specific ORSDS scores, as part of the construct validity assessment. A statistical tool for data reduction, it transforms a number of possibly correlated variables into a smaller number of uncorrelated variables called principal components. The first principal component accounts for as much of the variability in the data as possible, and each succeeding component accounts for as much of the remaining variability as possible. In all 6 situations, 4 to 5 highly correlated factors accounted for at least 70% of the variance (see Supplemental Digital Content 1, http://links.lww.com/AA/A276).

Known group validity refers to the prediction that patients reporting adverse symptoms should have higher symptom-specific ORSDS scores than patients without adverse symptoms. This was addressed for nausea, vomiting, and difficulty with urination. Known group validity was shown by the finding that nausea and vomiting validation measures correlated highly with relevant symptom-specific ORSDS scores. For example, symptom-specific scores for Peripheral (P < 0.0001), Neuraxial (P = 0.04), Regional (0.001), and GA (P = 0.001) were related to nausea or vomiting within the last 24 hours (Tables Q–T, see Supplemental Digital Content 2, http://links.lww.com/AA/A277).

Responsiveness refers to the ability of an instrument to detect changes once they have occurred and is evaluated by sensitivity, specificity, and κ statistics. Sensitivity is the proportion of true positives that are correctly identified. Specificity is the proportion of true negatives that are correctly identified. An optimal test would have 100% sensitivity and specificity. A κ statistic assesses the degree to which 2 measurements agree: with complete agreement κ = 1 and with random coincidence κ = 0. Responsiveness was determined using validation measures previously described,3 including questions about opioid-related symptoms and daily activities.

Dose-responsiveness. Correlation of composite and symptom-specific ORSDS scores with opioid intake (Table U, see Supplemental Digital Content 2, http://links.lww.com/AA/A277) examined the prediction that increased opioid intake would be associated with increased burden of opioid-related symptomatology. Significant correlation was found for the composite ORSDS scores for 4 of the 6 conditions, and also for some of the symptom-specific ORSDS scores.

DISCUSSION

Performance Characteristics

The composite ORSDS scores on POD 1 ranged from 0.19 (Peripheral) to 0.95 (GA) and followed the same trend as opioid consumption. Similar pain scores among groups on POD 1 (3.3–4.7) were associated with very different levels of opioid consumption (28–190 mg oral morphine equivalents, median). This indicates that achieving effective pain treatment in one clinical situation (posterior spine fusion) required 6.8 times as much opioid (after conversion to equivalents of oral morphine) as was needed for a less painful operation (ambulatory hand surgery). As expected, the outpatient surgical groups reported much less opioid use than did the inpatients.

Validation

The ORSDS was initially validated for patients undergoing outpatient laparoscopic cholecystectomy on POD 1.3 We describe validation of the scale on POD 1 for 2 orthopedic ambulatory patient populations and on PODs 1 and 3 for 2 orthopedic inpatient populations that continued to use opioids over this timeframe. In general, the ORSDS instrument fulfilled multiple predetermined validity criteria.

Internal consistency. High values of Cronbach α indicated high internal consistency and reliability; there was a high correlation among frequency, severity, bothersomeness, and summary scores for each of the 12 symptoms and also for composite ORSDS scores. The range reported by Apfelbaum et al.3 for Cronbach α was 0.92 to 0.98, similar to the values reported herein.

Content validity (extent of coverage of relevant concepts) was supported by previous validation of the ORSDS, and by demonstrating that as predicted, patients with CMEs generally had higher symptom-specific ORSDS scores than those without CMEs. This finding was very similar to that of Apfelbaum et al.3

Construct validity (whether the ORSDS correlates with measures of other variables in the hypothesized way) was addressed by examining correlation of ORSDS scores with other variables thought to be related to opioid intake, including activity levels, satisfaction, hours of help, satisfaction, and number of times vomited. For example, high composite ORSDS scores were determined to be significantly related to reduction in activity levels for 2 of the 4 groups: Peripheral patients on POD 1 and Regional patients on POD 3. Apfelbaum et al.3 also found a significant relationship between activity and ORSDS scores (for composite and some symptom-specific scores). Unlike Apfelbaum et al., we did not find a link between hours of assistance needed and composite ORSDS scores. Satisfaction was found by Apfelbaum et al. to be related to composite ORSDS scores3; this study found correlation between satisfaction and composite ORSDS scores for only 2 of 6 situations. Satisfaction is a broad concept with many determinants, and can reflect many aspects of the perioperative experience in addition to opioid-related side effects.

Exploratory principal components analysis found that 4 to 5 factors contributed approximately 70% of the variance, very similar to the finding by Apfelbaum et al.3 that 8 factors accounted for 84% of variance.

Known group validity refers to the prediction that patients reporting adverse symptoms should have higher symptom-specific ORSDS scores than patients without adverse symptoms. Symptom-specific ORSDS scores were consistently higher for patients in each group reporting adverse symptoms, measured dichotomously (any nausea or vomiting) or continuously (minutes of nausea), on both POD 1 and POD 3. This matches the findings by Apfelbaum et al.3

Responsiveness of the ORSDS instrument was demonstrated by favorable sensitivity, specificity, and the κ statistics. Responsiveness was shown for nausea or vomiting by correlating relevant symptom-specific ORSDS scores with direct assessment: specificity (nausea: 0.95–1), sensitivity (nausea: 0.29–0.88), and κ statistics (nausea: 0.80–0.86). Similar statistics were reported (for example) for nausea and CMEs by Apfelbaum et al.3: specificity 0.94, sensitivity 0.82, and κ 0.55.

Dose-responsiveness. Linear regression indicated that composite ORSDS scores were linearly related to opioid use for Peripheral, Regional, and GA patients, but not for Neuraxial patients. Patients may have titrated their analgesic usage to achieve acceptable pain scores, despite frequent occurrence of opioid-related side effects in groups with high opioid consumption. Previous validation of the ORSDS indicated that composite ORSDS scores were related to opioid use.12

Similarly, the number of CMEs is related to opioid intake. Peripheral patients received the least amount of opioids during the first 24 hours after surgery; consequently, they had the fewest reported CMEs. More than one-third (38%) of Peripheral patients reported no CMEs whereas an additional 18% of Peripheral patients reported only 1 CME (Table 4). Conversely, GA patients on POD 1 reported the highest amount of opioid intake. Predictably, 94% of GA patients reported ≥3 CMEs during the first POD.

Limitations

Symptoms measured by the ORSDS are potentially attributable to administration of opioids, but many also have other perioperative causes. Not all patients completed the study. The 5 inpatients who were unable to answer the questionnaire, as well as the 4 patients who withdrew consent, may be patients with uncaptured opiate-related side effects. The same can be conjectured for the 11 ambulatory patients who did not complete the questionnaire. Clinically, a variety of opioids were administered via different routes (oral, parenteral, and epidural) and then converted to oral equivalents to assess opioid dose responsiveness of the ORSDS. The accuracy of this conversion depends on data from other studies. Furthermore, different opioids and different routes of administration may have different efficacy and side-effect profiles.

There are numerous tools available to assess postoperative patient care. Severe postoperative sedation or confusion may be better measured by direct assessment.13 The Short Form 8 questionnaire provides adequate assessment for pain and physical function, but is deficient in assessing side effects and other comfort-related outcomes leading to the suggestion that the Short Form 8 could be coadministered with the ORSDS.14

The ORSDS provides a detailed and nuanced evaluation of side effects, and allows clinical investigators to focus on prominent or distressing side effects. One alternative, quantification of opioid utilization, is easy to measure but is not a patient-oriented outcome. Another alternative, evaluation of a single symptom such as nausea, ignores the importance of other side effects.2 It is possible that pharmacologic reduction of one side effect may worsen another (e.g., an antiemetic may reduce nausea at the expense of increased drowsiness or worsened concentration). Additionally, with the recent development of peripheral opioid antagonists, it is useful to note that fully half of the side effects are probably centrally mediated (difficulty concentrating, drowsiness or difficulty staying awake, feeling lightheaded or dizzy, feeling confused, feelings of general fatigue or weakness, headache). These are often prominent postoperative symptoms.

SUMMARY AND FUTURE DIRECTIONS

Performance analysis of the ORSDS indicated that the instrument provided information suitable for use as a primary or secondary outcome for clinical analgesic trials after the use of intraoperative GA or regional anesthesia, combined with a variety of postoperative analgesic therapies. Validation criteria were in general met. The ORSDS is a valid tool for assessment of opioid side effects after orthopedic surgery, and can be used for clinical trials across a variety of anesthetic and analgesic regimens.

RECUSE NOTE

Spencer S. Liu is section Editor of Pain Medicine for the Journal. This manuscript was handled by Franklin Dexter, section Editor of Economics, Education, and Policy, and Dr. Liu was not involved in any way with the editorial process or decision.

DISCLOSURES

Name: Jacques T. YaDeau, MD, PhD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Jacques T. YaDeau has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Name: Spencer S. Liu, MD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Spencer S. Liu has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Name: Matthew C. Rade, BA.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Matthew C. Rade has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Name: Dorothy Marcello, BA.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Dorothy Marcello has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Name: Gregory A. Liguori, MD.

Contribution: This author helped design the study, conduct the study, and write the manuscript.

Attestation: Gregory A. Liguori has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

ACKNOWLEDGMENTS

We thank Drs. Yan Ma and Heejung Bang for methodological and statistical consultations.

APPENDIX

ORSDS Questionnaire3
ORSDS Questionnaire3:
ORSDS Questionnaire3

REFERENCES

1. Aubrun F, Kalfon F, Mottet P, Bellanger A, Langeron O, Coriat P, Riou B. Adjunctive analgesia with intravenous propacetamol does not reduce morphine-related adverse effects. Br J Anaesth 2003;90:314–9
2. Portenoy RK, Thaler HT, Kornblith AB, Lepore JM, Friedlander-Klar H, Coyle N, Smart-Curley T, Kemeny N, Norton L, Hoskins W, Scher H. Symptom prevalence, characteristics and distress in a cancer population. Qual Life Res 1994;3:183–9
3. Apfelbaum JL, Gan TJ, Zhao S, Hanna DB, Chen C. Reliability and validity of the perioperative opioid-related symptom distress scale. Anesth Analg 2004;99:699–709
4. Portenoy RK, Thaler HT, Kornblith AB, Lepore JM, Friedlander-Klar H, Kiyasu E, Sobel K, Coyle N, Kemeny N, Norton L, Scher H. The Memorial Symptom Assessment Scale: an instrument for the evaluation of symptom prevalence, characteristics, and distress. Eur J Cancer 1994;30:1326–36
5. Myles PS, Hunt JO, Nightingale CE, Fletcher H, Beh T, Tanil D, Nagy A, Rubinstein A, Ponsford JL. Development and psychometric testing of a quality of recovery score after general anesthesia and surgery in adults. Anesth Analg 1999;88:83–90
6. Robins RW, Fraley RC, Kreuger RF Handbook of Research Methods in Personality Psychology. 1st ed. New York: Guilford Press, 2007
7. YaDeau JT, Cahill JB, Zawadsky MW, Sharrock NE, Bottner F, Morelli CM, Kahn RL, Sculco TP. The effects of femoral nerve blockade in conjunction with epidural analgesia after total knee arthroplasty. Anesth Analg 2005;101:891–5
8. Liu SS, Carpenter RL, Mulroy MF, Weissman RM, McGill TJ, Rupp SM, Allen HW. Intravenous versus epidural administration of hydromorphone: effects on analgesia and recovery after radical retropubic prostatectomy. Anesthesiology 1995;82:682–8
9. Wu CL. Acute postoperative pain. In: Miller RD ed. Miller's Anesthesia. 6th ed. Philadelphia: Elsevier Churchill Livingstone, 2005:2745–7
10. Gordon DB, Stevenson KK, Griffie J, Muchka S, Rapp C, Ford-Roberts K. Opioid equianalgesic calculations. J Palliat Med 1999;2:209–19
11. Amabile CM, Bowman BJ. Overview of oral modified-release opioid products for management of chronic pain. Ann Pharmacother 2006;40:1327–9
12. Zhao SZ, Chung F, Hanna DB, Raymundo AL, Cheung RY, Chen C. Dose-response relationship between opioid use and adverse effects after ambulatory surgery. J Pain Symptom Manage 2004;28:35–46
13. Inouye SK, Van Dyke CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI. Clarifying confusion: the confusion assessment method. Ann Intern Med 1990;113:941–8
14. Bost JE, Williams BA, Bottegal MT, Dang Q, Rubio DM. The 8-item Short-Form Health Survey and the physical comfort composite score of the quality of recovery 40-item scale provide the most responsive assessments of pain, physical function and mental function during the first 4 days after ambulatory knee surgery with regional anesthesia. Anesth Analg 2007;105:1693–700

Supplemental Digital Content

© 2011 International Anesthesia Research Society