If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author: Warren C. Hammert, MD, Department of Orthopaedics and Physical Performance, University of Rochester Medical Center, 601 Elmwood Ave., Box 665, Rochester, NY 14642.
We sought to determine whether subjective clinical improvement immediately after carpal tunnel release (CTR) was captured by Patient-Reported Outcomes Measurement Information System (PROMIS) Upper Extremity (UE), Physical Function (PF), Pain Interference (PI), and Depression.
Methods
Between September 2018 and January 2020, patients presenting to a single academic medical center hand clinic were asked to complete PROMIS UE, PF, PI, and Depression computer adaptive tests. In addition, patients who had CTR were asked to answer the following at their first postoperative clinic visit: “Since my last clinic visit, my condition is: (1) much better; (2) mildly better; (3) no change; (4) mildly worse; (5) much worse.” For each patient, the last clinic visit was the final preoperative visit. The PROMIS domain scores were compared before and after surgery using paired t tests. The percentage of patients subjectively reporting better symptoms was calculated.
Results
A total of 156 patients fit our inclusion criteria. The average number of days between the final preoperative visit and CTR was 7 (range, 0–30), and the average number of days between CTR and the first postoperative visit was 9 (range, 3–21). A total of 116 patients (74%) reported their carpal tunnel syndrome was better at their first postoperative visit. However, PROMIS UE, PF, and PI scores were significantly worse at the first postoperative visit, although not at clinically appreciable levels. There was no statistical or clinical difference in PROMIS Depression scores from pre- to postoperative time points.
Conclusions
Nearly 75% of patients subjectively report their carpal tunnel syndrome is better at their first follow-up visit within 3 weeks of CTR; however, PROMIS does not capture this improvement.
Clinical relevance
Hand surgeons evaluating patients shortly following CTR should be aware of the potential limitation of PROMIS to accurately capture immediate postoperative clinical outcomes. Disease-specific Patient-Reported Outcome Measures, such as the Boston Carpal Tunnel Questionnaire, may be preferred during this immediate postoperative timeframe.
Dr. Graham interviews Dr. Warren Hammert regarding his article "Evaluation of PROMIS Ability to Detect Immediate Postoperative Symptom Improvement Following Carpal Tunnel Release", which appears in the June 2021 issue of the Journal of Hand Surgery.
Carpal tunnel release (CTR) is the definitive treatment for carpal tunnel syndrome (CTS); as such, many studies have examined its efficacy in the short- and long-term.
understanding the recovery pattern of patients following CTR, including through the use of Patient-Reported Outcome Measures (PROMs), is critically important for stakeholders, such as hand surgeons.
One set of PROMs gaining popularity in hand and upper extremity care is the Patient-Reported Outcomes Measurement Information System (PROMIS).
Within the CTS literature, PROMIS domains have performed well compared with region- and condition-specific PROMs like the Michigan Hand Questionnaire and Boston Carpal Tunnel Questionnaire (BCTQ).
Further, PROMIS questionnaires have been analyzed in the setting of CTR to determine what change in scores represent the minimal clinically important difference (MCID).
Minimal clinically important differences for PROMIS Physical Function, Upper Extremity, and Pain Interference in carpal tunnel release using region- and condition-specific PROM tools.
Despite these promising studies, there remains a paucity of research examining whether PROMs in general, or more specifically the various PROMIS domains, capture the initial mild, but appreciable, postoperative (ie, within 21 days) symptom improvement that many patients anecdotally report to hand surgeons at the first postoperative follow-up visit. This concept refers to responsiveness of a PROM, which is the ability of a PROM to detect a change in the patient’s clinical condition or symptom being measured.
Understanding both the potential benefits and the disadvantages of PROMs, including PROMIS, is crucial as we transition to a value-based health care system, which considers outcomes achieved per dollar spent over the full cycle of care for a given condition.
In addition, such an appreciation of the benefits and limitations of PROMs—in both the short- and the long-term postoperative time periods—can also assist hand surgeons and patients in knowing when they can be used to appropriately monitor recovery and guide expectations. Thus, this study sought to determine whether PROMIS Upper Extremity (UE), Physical Function (PF), Pain Interference (PI), and Depression domains are able to capture immediate postoperative subjective clinical improvement after CTR. We hypothesized that PROMIS UE, PF, PI, and Depression questionnaires would not be able to capture immediate clinical improvement (ie, within 21 days) following CTR, as defined by estimates of MCID in the literature.
Minimal clinically important differences for PROMIS Physical Function, Upper Extremity, and Pain Interference in carpal tunnel release using region- and condition-specific PROM tools.
In addition, we hypothesized that these 4 PROMIS domains would not detect improvement immediately after CTR (ie, demonstrate responsiveness to the symptom[s] that may improve) at a rate that is detected via a single, subjective health status question.
Materials and Methods
Our institutional PROMIS data repository was queried for all patients diagnosed by a hand surgeon with CTS (International Classification of Diseases, 10th Revision: G56.0x) who subsequently underwent a CTR (Current Procedural Terminology Codes: 29848 and 64721) between September 2018 and January 2020. The majority of CTRs were open (n = 114; 73%) and were performed under local anesthesia in a procedure room. The remainder (n = 42; 27%) were endoscopic CTRs, which were performed under local anesthesia and sedation in our outpatient surgery center. In addition to the previously discussed diagnosis and procedure codes, our inclusion criteria also required that our sample included only adult patients (≥18 years) who had completed full sets of PROMIS UE, PF, PI, and Depression questionnaires at a final preoperative visit within 30 days of their CTR and a first postoperative visit up to 21 days after CTR. Patients who had a CTR in conjunction with other procedures were excluded.
Patient and PROMs data
For all patients, the following characteristics were collected: age (years), sex (female or male), self-reported race (white, black, or other), insurance type (commercial, Medicaid, Medicare, self-pay, workers’ compensation, or other), preoperative time point (ie, number of days between final preoperative clinic visit and surgery), and immediate postoperative time point (ie, number of days between surgery and the first postoperative clinic visit within 21 days).
In addition, PROMIS UE (v2.0), PF (v2.0), PI (v1.1), and Depression (v1.0) questionnaire results were collected at both the pre- and the postoperative appointments using tablet computers (Apple Inc., Cupertino, CA). All PROMIS questionnaires completed were computer adaptive tests collected as part of routine clinical care at our institution.
The PROMIS was developed through support from the U.S. National Institutes of Health to be normally distributed with a mean t score of 50 and an SD of 10.
Of note, higher scores for each PROMIS domain indicate more of the construct being measured (ie, higher PROMIS UE scores indicate greater upper extremity functional ability, and higher PROMIS Depression scores indicate higher levels of depressive symptoms).
In addition to completing the PROMIS questionnaires, patients were asked to answer the following health status question at their first postoperative clinic visit on a tablet computer (Apple Inc.): “Since my last clinic visit, my condition is: (1) much better; (2) mildly better; (3) no change; (4) mildly worse; (5) much worse.” This question was designed to allow patients a simplified way to easily share with their hand surgeon their improvement (or lack of improvement) in their hand condition. This health status question is similar to the Single Assessment Numeric Evaluation, which was shown to be a reasonable outcome measure of global hand function.
In each case, the last clinic visit was the final preoperative visit. This question aimed to evaluate each patient’s subjective feeling of symptom improvement or worsening shortly following CTR.
Statistical analysis
For continuous variables, means with SD or range were reported, whereas counts and percentages were used to report categorical variables. Answers on the subjective health status question were grouped into Better (representing responses of either (1) much better or (2) mildly better) or Not better (representing responses of (3) no change, (4) mildly worse, or (5) much worse). The percentage of patients reporting Better was reported. The difference in PROMIS domain scores from pre- to postoperative time points were also determined. The percentage of patients achieving a change in PROMIS scores that met or exceeded CTR MCID estimates was determined. Lastly, PROMIS scores (UE, PF, PI, and Depression) from the final preoperative visit were compared with those of the initial postoperative visit using paired t tests. This final analysis was done for all patients as well as the subgroup of patients who fell into the Better category.
For each PROMIS assessment, there is a range of MCID estimates in the setting of CTR based on the calculation method (eg, anchor-based method vs distribution-based method). In this study, the following MCID estimates were considered for PROMIS UE: 3.6 to 6.3
Minimal clinically important differences for PROMIS Physical Function, Upper Extremity, and Pain Interference in carpal tunnel release using region- and condition-specific PROM tools.
Minimal clinically important differences for PROMIS Physical Function, Upper Extremity, and Pain Interference in carpal tunnel release using region- and condition-specific PROM tools.
Minimal clinically important differences for PROMIS Physical Function, Upper Extremity, and Pain Interference in carpal tunnel release using region- and condition-specific PROM tools.
We were unable to identify PROMIS Depression MCID estimates in patients with CTS undergoing CTR in the literature. Using the current sample and the SD-based approach (ie, one-half of the SD of the PROM),
we estimated the PROMIS Depression MCID to be 4.6.
Assuming an effect size of 0.30, alpha of 0.05, and power of 80%, a sample size estimate based on a 2-tailed paired t test indicated that we would need at least 90 patients with pre- and post-CTR PROMIS scores. For all analyses, statistical significance was defined as a P value less than .05.
Our Institutional Review Board approved this study.
Results
A total of 156 patients fit our inclusion criteria (Table 1). The mean final preoperative visit was 7 days before CTR (range: 0–30 days), and the mean initial postoperative visit was 9 days after CTR (range: 3–21 days) (Table 1).
Of the 156 patients, 74% (n = 116) reported improvement (much better [n = 62] or mildly better [n = 54]) in CTS symptoms on the subjective health status question following CTR (Table 2). PROMIS UE (38.8 vs 35.1; P <.05), PF (43.6 vs 41.2; P < .05), and PI (56.9 vs 58.2; P < .05) scores were significantly worse at the initial postoperative visit (PROMIS scores listed as preoperative vs postoperative) (Table 3). Postoperative PROMIS Depression domain scores (46.0 vs 45.3; P = .17) were not significantly different from preoperative results (Table 3). When only evaluating patients that reported CTS symptom improvement following CTR, PROMIS UE (39.3 vs 36.2; P < .05) and PF (43.8 vs 41.9; P < .05) scores were significantly worse at the initial postoperative visit (PROMIS scores listed as preoperative vs postoperative) (Table 4). In contrast, PROMIS Depression (45.7 vs 44.0; P < .05) scores were significantly better at the initial postoperative visit. For this patient subgroup, there was no difference in PROMIS PI (56.8 vs 57.5; P = .24) scores from the final pre- to the immediate postoperative clinic visit (Table 4).
Table 2Subjective Health Status From Final Preoperative to First Postoperative Visit
Among all patients, change in PROMIS UE met only 1 MCID estimate (ie, 3.6)13; however, in our sample, it actually demonstrated a clinically relevant worsening in symptoms at the first postoperative follow-up visit, which is expected due to surgical pain and initial limitations. Similarly, change in PROMIS PF met only 1 MCID estimate (ie, 1.8),
Minimal clinically important differences for PROMIS Physical Function, Upper Extremity, and Pain Interference in carpal tunnel release using region- and condition-specific PROM tools.
and it also demonstrated a worsening in symptom severity at first postoperative follow-up. This finding was also not surprising. Neither changes in PROMIS PI nor Depression met any MCID estimate in our full patient sample.
When evaluating only the patient subgroup who reported subjective improvement in symptoms (ie, Better group), only the change in PROMIS PF met an MCID estimate (ie, 1.8)12; however, as with the analysis with the full patient sample, this change demonstrated a worsening in symptom severity at the first postoperative visit. Changes in PROMIS UE, PI, and Depression did not exceed any MCID estimate in our patient subgroup that reported subjective clinical improvement at short-term postoperative follow-up.
The percentage of patients who demonstrated clinical improvement as measured by PROMIS UE (Fig. 1), PF (Fig. 2), PI (Fig. 3), and Depression (Fig. 4) assessments was significantly lower than the percentage of patients who reported symptom improvement using the health status question (P < .05 for all comparisons).
Figure 1The percentage of patients achieving appreciable clinical improvement based on different MCID estimates for PROMIS UE in the literature. The percentage of patients achieving appreciable clinical improvement based on the health status question is also shown; this percentage is significantly higher than the percentage achieving appreciable clinical improvement based on any MCID estimate (P < .05). MHQ, Michigan Hand Questionnaire.
Figure 2The percentage of patients achieving appreciable clinical improvement based on different MCID estimates for PROMIS PF in the literature. The percentage of patients achieving appreciable clinical improvement based on the health status question is also shown; this percentage is significantly higher than the percentage achieving appreciable clinical improvement based on any MCID estimate (P < .05). MHQ, Michigan Hand Questionnaire.
Figure 3The percentage of patients achieving appreciable clinical improvement based on different MCID estimates for PROMIS PI in the literature. The percentage of patients achieving appreciable clinical improvement based on the health status question is also shown; this percentage is significantly higher than the percentage achieving appreciable clinical improvement based on any MCID estimate (P < .05). MHQ, Michigan Hand Questionnaire.
Figure 4The percentage of patients achieving appreciable clinical improvement based on our distribution-based MCID estimate for PROMIS Depression. The percentage of patients achieving appreciable clinical improvement based on the health status question is also shown; this percentage is significantly higher than the percentage achieving appreciable clinical improvement based on our MCID estimate (P < .05).
The use of PROMs, such as PROMIS, has become a common element of care delivery, especially as health care systems transition toward rewarding value. In general, PROMIS and other instruments are better for understanding groups of patients and trends throughout an episode of care, as opposed to determining whether a single patient did or did not clinically improve following an intervention. However, when used appropriately, PROMIS can also play an important role helping hand surgeons stratify risk, better set expectations broadly, and monitor recovery compared with what is typically expected in similar patient populations following treatment or surgery. For example, a patient with higher a PROMIS Depression score may be expected to have more difficulty recovering from surgery and less symptom relief. Understanding this prior to surgery can help guide preoperative discussions in a more meaningful way. Anecdotal evidence suggests that patients appreciate some level of clinical improvement immediately after surgery at their first follow-up visit (ie, within 21 days of CTR). In the present study, our main finding was that nearly three-quarters of patients reported subjective clinical improvement within 21 days of undergoing a CTR; however, PROMIS UE, PF, PI, and Depression computer adaptive tests were unable to capture this immediate postoperative subjective symptom improvement. In addition, when evaluating changes in PROMIS scores, we found evidence of appreciable worsening of functional status, which we believe may be secondary to patients being more cautious in using their hand immediately following surgery. Therefore, it can be concluded that the studied PROMIS domains are not responsive to early postoperative symptom improvement following CTR; however, the PROMIS domain scores are likely accurate and more reflective of surgical morbidity and the impact of CTR itself, whereas the subjective clinical improvement reported by patients is indicative of symptom improvement (eg, nerve-related symptoms), but is not specifically evaluated by the PROMIS domains studied.
Similar findings were also noted in an analysis of 44 patients by Kazmers et al,
which showed that PROMIS UE, PF, and PI did not significantly improve to a clinically appreciable level following CTR at about 15 days after surgery; this was despite 89% of patients reporting at least a minimal amount of appreciable improvement. Importantly, in the same study, the shortened Disabilities of the Arm, Shoulder, and Hand (QuickDASH), another commonly used PROM that is region-specific (ie, upper extremity), also did not capture this early improvement.
demonstrated that the average change in QuickDASH scores at the 2-week postoperative follow-up time point was 15, 10, and 17 points for mild, moderate, and severe CTS, as measured by electrodiagnostic studies. However, the statistical improvement (ie, P < .05) in symptoms—as measured by PROMs—is appreciated at the 3-month postoperative follow-up.
suggesting that QuickDASH may not always capture clinical improvement.
The PROMIS, as well as other general or non-disease–specific PROMs, may not be able to capture changes in all symptoms for each condition. For example, a key symptom of CTS improved by a CTR is nighttime awakening (NTA); this is one of the first symptoms to be relieved after CTR.
Lander S, Lander A, Hammert WC. Outcomes of patients with and without nighttime awakening symptoms after carpal tunnel release [published online ahead of print February 26, 2020]. Hand (N Y). https://doi.org/10.1177/1558944719895620.
Improvement in NTA symptoms is not captured in the PROMIS questionnaires studied because no questions specifically ask about NTA or sleep, and NTA does not necessarily have an impact on function or daytime pain. Importantly, however, other PROMIS domains, such as PROMIS Sleep Disturbance,
may be better at evaluating NTA. Further, many patients may also find that their hand paresthesia, which was bothersome but never caused pain or functional limitations, improves shortly following a CTR as well. The PROMIS assessments may not capture such symptom improvement. Although the DASH and QuickDASH includes 2 questions that could capture the impact of NTA (ie, the severity of tingling [pins and needles] in the arm, shoulder, or hand question and the difficulty with sleeping question), they are not a major portion of these scales. Further, they do not appear to be responsive enough to capture immediate postoperative CTR improvement either. Further, PROMIS domains have been shown to perform similarly to many hand and upper extremity PROMs.
Given the early improvement following CTR owing to resolution of paresthesia or NTA, a condition-specific outcome instrument that measures disease activity, such as BCTQ, may be the only scale able to capture this early change. Hand surgeons, payors, and policy makers should take our findings into consideration when monitoring CTR recovery and designing alternative payment models that consider these specific PROMs as a measure of quality.
Our study had limitations. First, our findings are derived from a single institution, so the generalizability of our work is unknown. Second, we utilized PROMIS UE, PF, PI, and Depression as the set of PROMs; therefore, our findings may not be consistent if other, more hand and upper extremity–specific PROMs were evaluated. However, prior research suggests that our findings may be consistent across other institutions and PROMs.
Third, our health status question does not ask the reason why patients feel like they have or have not improved from CTR. We do not know the exact reason why nearly three-quarters of patients in our sample report clinical improvement. Future research that specifically asks the underlying reason behind each patient’s general feeling of clinical improvement would be of value. The PROMIS domains have been shown to be responsive to patients undergoing CTR.
Therefore, we believe our findings more represent a question of the timing of when the PROMIS domains were administered. It is plausible that our findings represent the competition between the effects of the surgery itself and the symptom improvement provided by the intervention. It is also possible that the feeling of clinical improvement noted by many patients via the health status question is a cognitive bias in which patients equate surgical intervention with getting better, despite that not necessarily being the case. Many patients do, in fact, express a preference for surgical management over nonsurgical intervention, and therefore, they may assume surgical management fixes their condition.
Fourth, we grouped all patients who responded (1) much better or (2) mildly better into a single category. However, we believe this is appropriate because our study aimed to compare the percentage of patients subjectively noting improvement with the percentage of patients achieving a change in PROMIS scores equivalent to the MCID. Finally, we only included patients with preoperative PROMIS scores within a month of surgery. We used this time point to minimize the chance of other conditions affecting PROMIS scores between the preoperative and the initial postoperative appointment; we considered this to be important because PROMIS domains are not specific to hand conditions.
Overall, we found that nearly three-quarters of patients with CTS report clinical improvement immediately following CTR (ie, within 21 days of surgery). In contrast, PROMIS UE, PF, PI, and Depression domains were not able to detect this immediate clinical improvement following CTR. Whereas the PROMIS scores collected at this early postoperative follow-up time point are probably accurate, the surgical morbidity and symptoms related to the CTR itself may be the reason for worsening or stable scores. Patients likely report clinical improvement because of the immediate benefit of treatment on symptoms for which CTR is often offered, but which are not captured in the PROMIS domains studied (eg, NTA). This highlights the importance of longer follow-up after CTR. Hand surgeons should be aware that they should not expect notable improvements in PROMIS scores immediately following CTR, but that patients may still be on a positive recovery trajectory. Such information can be shared with patients as well and be part of the setting of expectations. Further, health policy and payor stakeholders should acknowledge the limitations of PROMs, as demonstrated in this study, when designing alternative payment models that reward quality. Future research to determine the postoperative time points at which PROMIS UE, PF, PI, and Depression domains are able to capture clinically appreciable improvement in CTS symptoms after CTR is needed. However, based on the current study’s findings, PROMIS scales may be best suited at evaluating clinical outcomes post-CTR after the full treatment effect is thought to have occurred, whereas more disease-specific PROMs, such as the BCTQ, may be important instruments in evaluating the immediate effects of treatment.
References
Tang C.Q.Y.
Lai S.W.H.
Tay S.C.
Long-term outcome of carpal tunnel release surgery in patients with severe carpal tunnel syndrome.
Minimal clinically important differences for PROMIS Physical Function, Upper Extremity, and Pain Interference in carpal tunnel release using region- and condition-specific PROM tools.
Lander S, Lander A, Hammert WC. Outcomes of patients with and without nighttime awakening symptoms after carpal tunnel release [published online ahead of print February 26, 2020]. Hand (N Y). https://doi.org/10.1177/1558944719895620.
00102-7&id=gr1.jpg){kind=link}
00102-7&id=gr2.jpg){kind=link}
00102-7&id=gr3.jpg){kind=link}
00102-7&id=gr4.jpg){kind=link}