Patient-reported symptom interference as a measure of postsurgery functional recovery in lung cancer

Abstract

Context

Few empirical studies have combined the patient’s perspective (patient-reported outcomes, or PROs) with clinical outcomes (risk for complications, length of hospital stay, return to planned treatment) to assess the effectiveness of treatment after thoracic surgery for early-stage non-small cell lung cancer (NSCLC).

Objectives

Quantitatively measure PROs to assess functional recovery postsurgery.

Methods

Treatment-naïve patients (N=72) with NSCLC who underwent either open thoracotomy or video-assisted thoracoscopic surgery (VATS) used the MD Anderson Symptom Inventory (MDASI) to report symptom interference with general activity, work, walking, mood, relations with others, and enjoyment of life for 3 months postsurgery. Functional recovery was defined as interference scores returning to presurgery levels. The MDASI’s sensitivity to change in functional recovery over time was evaluated via its ability to distinguish between surgical techniques.

Results

Interference scores increased sharply by day 3 postsurgery (all P<0.001), then returned to baseline levels via different trajectories. Patients who had unscheduled clinic visits during the study period reported higher scores on all 6 MDASI interference items (all P<0.05). Compared with the open-thoracotomy group, the VATS group returned more quickly to baseline interference levels for walking (18 vs 43 days), mood (8 vs 19 days), relations with others (4 vs 16 days) and enjoyment of life (15 vs 41 days) (all P<0.05).

Conclusion

Repeated measurement of MDASI interference characterized functional recovery after thoracic surgery for NSCLC and was sensitive to VATS’ ability to enhance postoperative recovery. Further study of the clinical applicability of measuring recovery outcomes using PRO-based functional assessment is warranted.

INTRODUCTION

The symptomatic sequelae of major thoracic surgery and its complications may have a significant negative impact on functional recovery for patients with cancer (1–3). Because time to recovery dictates how quickly a patient can attain normal physical functioning and subsequent planned cancer therapies can begin, if needed, understanding functional recovery is particularly important. Delayed or cancelled adjuvant therapy can adversely affect long-term clinical outcomes (4); therefore, measuring the severity of both symptoms and functional impairment from surgery should be a routine component of perioperative care.

Functional recovery after surgery is a complex process that involves physical, psychological, and social domains and that is influenced by various factors, including patient, surgical, and anesthetic characteristics and the presence of adverse events (5,6). In general, functional status after major surgery is characterized by a period of immediate deterioration followed by gradual rehabilitation to baseline level (6), yet actual functional recovery does not always align with metrics typically used to assess recovery, such as length of hospital stay (LOS). For example, although the median LOS for patients with non-small cell lung cancer (NSCLC) has been reported to be 6 days (interquartile range, 4–9 days) (7), the true postoperative recovery period for returning to preoperative levels of usual activity varies, ranging from 1–3 months (8,9). The rapid adoption of enhanced recovery pathways also is influencing functional recovery after surgery, in terms of both LOS and postdischarge functioning.

To grasp the potential impact of newly developed enhanced recovery and operative techniques, one must first be able to understand the effects of such techniques on a reasonable outcome, such as change in functional status over time. However, there is lack of agreement about a standard outcome measure to assess functional recovery after major surgery or hospital discharge. Patient-reported outcomes (PROs) are not typically collected during postoperative care (except for pain scores). Nonetheless, functional status can be reported by patients using short PRO questionnaires and electronic data-capture methods, even after discharge. PROs have been widely accepted in clinical research (10), are increasingly used in routine oncology practice (11), and have been endorsed by the US Food and Drug Administration for use in drug labeling-claim trials (12).

The introduction of patient-reported functional outcomes in patient care is relatively novel in current perioperative practice, primarily because of concerns about the reliability and comparability of PRO data and methodological issues related to how PROs can be deployed and understood in clinical studies and practice (13). Few studies with PRO-based functional status as a primary or secondary outcome have longitudinally described functional recovery trajectories throughout the perioperative period, and few empirical studies have combined the patient’s perspective (PROs) with clinical outcomes (risk for complications, length of hospital stay, return to planned treatment) to assess treatment effectiveness (14,15). Integrated into postoperative care, such PRO reports could provide additional information to clinicians about recovery and could be an additional metric for comparing different surgical approaches, with minimal increase in cost.

The objective of the current study was, through a longitudinal PRO-based investigation, to establish a method for monitoring functional recovery in patients with early stage NSCLC undergoing either open thoracotomy or video-assisted thoracoscopic surgery (VATS), a less-invasive procedure that is expected to favor more-rapid postoperative recovery (16–18). We have previously reported on the utility of symptom-severity measurement from repeated administration of the MD Anderson Symptom Inventory (MDASI) after thoracic surgery (19). Here, we hypothesized that MDASI interference, as a measure of daily functioning, would be sensitive to differences in the recovery trajectory related to surgery type (open thoracotomy vs. VATS).

METHODS

Patients

We prospectively recruited treatment-naïve patients with stage I or II NSCLC who were scheduled for thoracic surgery (either open thoracotomy or VATS) at The University of Texas MD Anderson Cancer Center in Houston, Texas. Eligible patients were required to be at least 18 years old, to understand English and to be able to understand the study requirements, and to be willing and able to respond to repeated MDASI assessments administered via a computerized, telephone-based interactive voice response (IVR) system after hospital discharge. The study was approved by the MD Anderson Institutional Review Board, and all participants gave written informed consent.

Study design

Outcome measures

Functional status was assessed with the MDASI, a psychometrically validated PRO assessment tool that measures multiple cancer-related symptoms and their interference with daily functioning (20). The MDASI includes not only 13 symptom items, but also 6 items that measure symptom interference with daily functioning in response to the question, “How have your symptoms interfered with your life?” The interference items include physical aspects (general activity, work, walking) and psychological aspects (mood, relations with others, enjoyment of life). Each interference item is rated on an 11-point scale, with 0 = “did not interfere” and 10 = “interfered completely” in the previous 24 hours. MDASI interference was the outcome measure for this study.

While hospitalized, patients completed a paper-and-pencil version of the MDASI at the time of enrollment (presurgery baseline) and on day 3 postsurgery. Patients were given a demonstration of the IVR system before they were discharged. Beginning on day 7 postsurgery and continuing to 3 months postsurgery or until the patient started other cancer treatments, the IVR system called patients once each week.

At baseline, the SF-12 (21) was administered to measure presurgery functional health and well-being. Eastern Cooperative Oncology Group performance status (ECOG PS) ratings (22), demographic characteristics, postoperative complications, and other clinical variables also were recorded.

Statistical analysis

Longitudinal analyses of all available data were conducted. To be included in the analysis, a patient must have provided MDASI data at baseline and at least 2 additional time points during follow-up. Presurgery interference levels and certain perioperative factors and were compared by surgery type, using the Chi-square test (for categorical data) or Wilcoxon rank sum test (for continuous data). To describe symptomatic recovery on the basis of repeated assessments of the MDASI interference items over time, we used mixed-effects models to compare the MDASI interference scores between baseline and other critical postsurgery time points: week 1, month 1, month 2, and month 3. Medians, means, interquartile ranges, standard deviations, and 95% confidence intervals are reported.

To identify risk factors for higher interference levels, we used mixed-effect models with multiple candidates from demographic and clinical factors, including age, sex, race, marital status, baseline ECOG PS, baseline SF-12 physical composite score and mental component score, type of surgery, length of hospitalization, estimated blood loss, postsurgery pulmonary complications and cardiovascular complications, and unscheduled clinic visits during the 3 months postsurgery. In these models, the outcome measure was the score of each interference item as a continuous variable; the time variable (days from surgery) was considered as continuous, divided into 2 segments separated by day 3, when the highest scores were reported.

We defined “postoperative functional recovery” for an interference item as the patient having reported 2 contiguous postoperative interference ratings at or below the preoperative (baseline) level. Recovery days by surgery type were estimated for each interference item using Kaplan–Meier analysis, and log-rank tests were used to compare time to recovery by surgery type. All significance levels were set as 2-sided with alpha = 0.05. SAS 9.3 (SAS Institute, Inc., Cary, NC, USA) was used for all analyses.

RESULTS

Of the 77 patients with early stage NSCLC initially recruited for the study, 72 provided interference data at baseline and at ≥ 2 additional time points, and were thus included in the analysis. Table 1 demonstrates the distribution of patient demographic and clinical factors by surgical procedure (open thoracotomy vs VATS lobectomy). No significant difference between surgery types was found for any factor. Presurgery median scores for all MDASI interference items were 0 (63%–81% of patients reported 0), and no significant differences between patients were found for any items by surgery type.

TABLE 1.

Demographic and clinical characteristics

	Open thoracotomy (n = 40)	VATS lobectomy (n = 32)	P
	Mean (SD)	Mean (SD)
Age, years	66.4 (11.0)	65.2 (9.6)	0.64
Presurgery SF-12 physical component score	49.8 (9.0)	50.4 (8.8)	0.75
Presurgery SF-12 mental component score	50.9 (8.1)	49.4 (10.9)	0.51
Length of stay, days	6.3 (4.3)	6.2 (3.4)	0.86
Estimated blood loss, mL	385.2 (695.6)	257.2 (550.4)	0.40
	n (%)	n (%)	P
Age			0.86
< 60 years	12 (30)	9 (28)
≥ 60 years	28 (70)	23 (72)
Sex			0.25
Male	23 (58)	14 (44)
Female	17 (42)	18 (56)
Race			0.97
Non-Hispanic white	38 (95)	27 (84)
Other	2 (5)	5 (16)
Marital status			0.74
Married	26 (65)	14 (44)
All others	14 (35)	18 (56)
Preoperative ECOG PS			0.76
Good (0–1)	34 (85)	28 (88)
Poor (2–4)	6 (15)	4 (12)
Smoked ≥100 cigarettes in your lifetime			0.32
Yes	33 (83)	29 (91)
No	7 (17)	3 (9)
Comorbid conditions (Charlson Index)			0.15
No (0)	5 (13)	1 (3)
Yes (≥ 1)	35 (87)	31 (97)
Stage			0.12
I	29 (72)	28 (87)
II	11 (28)	4 (13)
Length of hospital stay			0.91
≤ 6 days	28 (70)	22 (69)
7 days or more	12 (30)	10 (31)
Any complications			0.59
No	25 (62)	18 (56)
Yes	15 (38)	14 (44)
Pulmonary complications			0.39
No	30 (75)	21 (66)
Yes	10 (25)	11 (34)
Cardiovascular complications			0.67
No	34 (85)	26 (81)
Yes	6 (15)	6 (19)
Unscheduled postoperative clinic visit			0.27
No	31 (78)	28 (88)
Yes	9 (22)	4 (12)
Adjuvant treatment within 3 months post surgery	3 (8)	2 (6)	0.836
Requested drop-off before month 3	7 (15)	6 (19)	0.891

ECOG PS, Eastern Cooperative Oncology Group performance status; VATS, video-assisted thoracoscopic surgery.

Longitudinal profiles of functional status

Of the 72 patients included in the analysis, 56 (77%) completed MDASI data for the entire 3-month study; 16 patients withdrew during the follow-up period (4 in month 1, 7 in month 2, and 5 after month 2), either because they began receiving other cancer therapies (n=5) or because they elected to withdraw (n=11). The compliance rate as assessed at major time points ranged from 80%–98% for the open surgery group and from 75%–91% for the VATS group. We observed no significant differences in compliance by type of surgery at any time point.

Fig. 1 shows that mean MDASI interference scores increased rapidly after surgery, peaking at day 3, and then returned to preoperative levels over time. Mean interference scores were no higher than their presurgery levels by month 2 postsurgery for patients who underwent VATS lobectomy and by month 3 postsurgery for patients who underwent open thoracotomy (Fig. 1). Interference ratings did not vary significantly by surgery type at baseline or at the 1-month timepoint; however, the VATS group reported significantly less interference with general activity and work at week 1, month 2, and month 3, compared with the open thoracotomy group (all P < 0.05). Similarly, the VATS group reported significantly less interference with mood at month 2, with relations with others at week 1 and month 3, and with enjoyment of life at month 2.

FIGURE 1. Change in average MDASI interference ratings during the 3 months postsurgery, by surgery type.

MDASI, MD Anderson Symptom Inventory; VATS, video-assisted thoracoscopic surgery.

Dynamic change of functioning recovery in the level of postoperative interference outcomes was examined by mixed-effects modeling (Table 2). Consistently for all interference items, scores increased from the day of surgery to day 3 (all P < 0.001), and then decreased between day 3 and month 3 (P < 0.001). Table 2 also presents mixed-effect modeling of risk factors associated with postoperative interference outcomes. Over the 3 months postsurgery, scores for all MDASI interference items were significantly lower for patients who underwent VATS lobectomy than for patients who underwent open lobectomy.

TABLE 2.

Factors Associated with Poor MDASI Interference Ratings Over Time (N = 72, Number Of Observations = 789)

MDASI interference item		Estimate (SE)	P
General activity
Change over time	Days from surgery (0–3)	0.24 (0.03)	< 0.001
	Days from surgery (3–100)	–0.29 (0.03)	< 0.001
Risk factors
Age		–0.03 (0.01)	0.01
ECOG PS (≥ 1 vs. 0)		0.63 (0.26)	0.02
Comorbidity (yes vs. no)		1.28 (0.38)	< 0.001
Open thoracotomy vs. VATS lobectomy		0.99 (0.22)	< 0.001
Days of hospitalization (≥ 7 vs. < 7)		1.00 (0.26)	< 0.001
Pulmonary complications (yes vs. no)		0.72 (0.23)	0.002
Unscheduled clinic visit within 3 months postsurgery (yes vs. no)		0.33 (0.16)	0.04
Work
Change over time	Days from surgery (0–3)	0.29 (0.03)	< 0.001
	Days from surgery (3–100)	–0.34 (0.04)	< 0.001
Risk factors
Age		–0.04 (0.01)	0.01
ECOG PS (≥ 1 vs. 0)		0.68 (0.29)	0.02
Comorbidity (yes vs. no)		1.33 (0.41)	0.001
Open thoracotomy vs. VATS lobectomy		1.73 (0.25)	< 0.001
Days of hospitalization (≥ 7 vs. < 7)		0.94 (0.28)	< 0.001
Pulmonary complications (yes vs. no)		0.77 (0.25)	0.004
Unscheduled clinic visit within 3 months postsurgery (yes vs. no)		0.57 (0.17)	0.001
Walking
Change over time	D a y s from surgery (0–3)	0.11 (0.03)	< 0.001
	Days from surgery (3–100)	–0.13 (0.03)	< 0.001
Risk factors
Open thoracotomy vs. VATS lobectomy		0.48 (0.19)	0.01
Baseline SF-12 physical composite score		–0.05 (0.01)	< 0.001
Pulmonary complications (yes vs. no)		0.51 (0.20)	0.014
Unscheduled clinic visit within 3 months postsurgery (yes vs. no)		0.95 (0.14)	< 0.0001
Mood
Change over time	D a y s from surgery (0–3)	0.10 (0.02)	< 0.001
	Days from surgery (3–100)	–0.12 (0.03)	< 0.001
Risk factors
Age		–0.03 (0.01)	0.001
Baseline SF-12 mental composite score		–0.04 (0.01)	0.0002
Comorbidity (yes vs. no)		0.85 (0.30)	0.005
Stage (II vs. I)		0.49 (0.20)	0.015
Open thoracotomy vs. VATS lobectomy		0.68 (0.18)	0.0002
Days of hospitalization (≥ 7 vs. < 7)		0.48 (0.20)	0.018
Pulmonary complications (yes vs. no)		0.69 (0.19)	0.0002
Unscheduled clinic visit within 3 months postsurgery (yes vs. no)		0.48 (0.12)	0.0002
Relations with others
Change over time	D a y s from surgery (0–3)	0.07 (0.02)	0.004
	Days from surgery (3–100)	–0.08 (0.02)	< 0.001
Risk factors
Age		–0.02 (0.008)	0.03
Baseline SF-12 mental composite score		–0.03 (0.01)	0.003
Stage (II vs. I)		0.48 (0.19)	0.011
Open thoracotomy vs. VATS lobectomy		0.57 (0.16)	< 0.001
Days of hospitalization (≥ 7 vs. < 7)		0.40 (0.18)	0.028
Unscheduled clinic visit within 3 months postsurgery (yes vs. no)		0.38 (0.11)	0.0009
Enjoyment of life
Change over time	D a y s from surgery (0–3)	0.14 (0.03)	< 0.001
	Days from surgery (3–100)	–0.17 (0.04)	< 0.001
Risk factors
Baseline SF-12 mental composite score		–0.04 (0.01)	0.003
Comorbidity (yes vs. no)		1.27 (0.36)	0.0004
Open thoracotomy vs. VATS lobectomy		0.93 (0.22)	< 0.0001
Days of hospitalization (≥ 7 vs. < 7)		0.58 (0.25)	0.019
Pulmonary complications (yes vs. no)		0. 69 (0.22)	0.002
Unscheduled clinic visit within 3 months postsurgery (yes vs. no)		0.43 (0.15)	0.005

ECOG PS, Eastern Cooperative Oncology Group performance status; MDASI, MD Anderson Symptom Inventory; SE, standard error; VATS, video-assisted thoracoscopic surgery.

Presurgery risk factors for more-severe symptom interference included: younger age for all interference items except enjoyment of life; presence of comorbidities for general activity, work, mood, and enjoyment of life; ECOG-PS ≥ 1 for general activity and work; lower (indicating poorer) preoperative SF-12 mental component score for mood, relations with others, and enjoyment of life; stage II lung cancer for mood and relations with others; and a lower preoperative SF-12 physical component score for walking.

Postsurgery risk factors for higher interference scores were: unscheduled clinic visits, for all interference items; LOS ≥ 7 days, for all interference items except walking; and pulmonary complications, for all interference items except relations with others.

Time to functional recovery, measured by MDASI interference

Table 3 shows that interference with general activity and work returned to presurgery levels for all patients within approximately 2 months. Walking and enjoyment of life took 1 month to recover, interference with mood took 2 weeks, and relations with others took only 1 week to return to its presurgery level.

TABLE 3.

Kaplan–Meier Estimated Functional Recovery Time (Days from Surgery)

MDASI interference item	Median days to recovery (95% confidence interval)a			P
	Overall	Open thoracotomy	VATS lobectomy
General activity	63 (36–89)	57 (29–84)	63 (17–109)	0.979
Work	63 (29–96)	–	46 (24–68)	0.064
Walking	29 (16–42)	43 (23–63)	18 (3–34)	0.030
Mood	14 (5–23)	19 (5–33)	8 (3–12)	0.016
Relations with others	8 (3–13)	16 (5–27)	4 (3–6)	< 0.0001
Enjoyment of life	30 (19–41)	41 (12–70)	15 (3–36)	0.020

MDASI, MD Anderson Symptom Inventory; VATS, video-assisted thoracoscopic surgery.

^aFunctional recovery for an interference item was defined as the patient having reported 2 contiguous interference ratings at or below the preoperative (baseline) level.

Confirming the sensitivity of MDASI interference for indicating recovery after thoracic surgery, Fig. 2 presents the profiles of and significant differences in functional recovery between VATS lobectomy and open thoracotomy, by Kaplan–Meier analysis. Compared with patients who underwent open thoracotomy, patients who underwent VATS lobectomy took significantly less time to return to baseline levels for interference with walking (18 vs 43 days, P = 0.03), mood (8 vs 19 days, P = 0.02), relations with others (4 vs 16 days, P < 0.001) and enjoyment of life (15 vs 41 days, P = 0.02) (Table 3). No differences by surgery type were found for interference with general activity and work.

FIGURE 2. Functional recovery to preoperative status, by surgery type.

VATS, video-assisted thoracoscopic surgery. *P by log-rank test.

DISCUSSION

This longitudinal study demonstrated the potential utility of the patient-reported MDASI interference items for measuring functional recovery outcomes after thoracic surgery and for detecting differences in expected return of functioning by type of procedure (here, VATS lobectomy vs standard open thoracotomy). To our knowledge, this is the first study to define postoperative functional recovery with a PRO-based assessment as an outcome measure (18,23).

Real-time reporting of both physical and psychological functional status to health care providers, which is possible via the MDASI, may improve the likelihood of effective evaluation of postoperative recovery and, as a result, improve patient care (2). The MDASI is a flexible, easily completed, psychometrically valid assessment tool that can be deployed in various ways, including paper and pencil, IVR, and other electronic data capture methods. For example, MDASI information can be collected using a web interface, such as a patient portal in an electronic health record system, that can be deployed in several different ways, including smartphones, tablets, or home computers. The MDASI can be completed in less than 5 minutes with any of these modes. We have shown that postdischarge symptom burden was easily captured by an IVR system presenting MDASI symptom-severity items to patients at home (2). We have also demonstrated the MDASI’s sensitivity to important differences in symptom severity by the type of procedures presented here (open thoracotomy vs VATS) (19).

As expected, greater functional impairment was experienced by patients after open thoracotomy surgery than after VATS lobectomy (24,25). This may be related to a combination of insults associated with open thoracotomy, such as rib retraction, resection, or fracture, costovertebral joint dislocation, intercostal nerve injury, and/or irritation of the pleura by chest tubes (26). Our finding of differences in patient-reported interference between VATS lobectomy and standard open thoracotomy mirrors previously reported clinical benefits from thoracoscopic lobectomy for early-stage NSCLC (27,28). The “interference with walking” item was especially sensitive for capturing the differences in physical functioning by procedure. Collectively, these results demonstrate that the MDASI is sensitive enough to detect differences in postoperative functional status by type of procedure, and that MDASI results might be used as an outcome metric for comparing other procedural differences in the delivery of perioperative care.

Patients who had an unscheduled clinic visit during the 3 months postsurgery had significantly higher interference scores than those who did not, and unplanned visits are frequently associated with poorly managed pain and other symptoms (29). In clinical trials, unscheduled clinic visits have been used as an outcome metric for the efficacy and quality of perioperative care techniques, such as telemedicine follow-up (30) and enhanced recovery practices (31). The association between this clinical outcome (prolonged hospitalization, unscheduled visits, and complications), more comorbidities, poorer performance status, and patient-reported MDASI interference supports the reliability of PRO-based functional measures in evaluating symptom-induced functional impairment after lung cancer surgery.

Using MDASI interference ratings as an indicator of patient-reported functional status, we not only defined the severity of functional interference in this study, but also defined the time course of postoperative recovery after thoracic surgery for NSCLC. The late phase of postoperative recovery has been described with a function-based definition: the time “from hospital discharge to return to usual function and activities” (6). This definition could be useful for evaluating perioperative care and procedures, given the lack of agreement in contemporary clinical practice on what constitutes optimal functional recovery. In the current study, repeated measurements of MDASI interference at selected critical timepoints, beginning with a preoperative assessment, sufficiently and sensitively captured significant differences in the time course of functional impairment and the timing of return to preoperative function for 2 different surgery types for early-stage NSCLC. Our previous data showed that the distribution-based minimal clinically important difference (0.5 SD) is 1.3. In the current data, the difference between month 2 and baseline for walking was 1.6 in the open thoracotomy group, suggesting a clinically meaningful difference.

Implied in this is that quantification of recovery requires measurement at baseline and comparison with an established course from population data (6). Considering that the average LOS is less than 1 week, understanding the time course of postdischarge functional recovery is vital not only for informing treatment teams tasked with reducing surgical morbidity, but also for characterizing patient status once the patient has returned to his or her intended cancer care plan as an outpatient in the community healthcare setting (4).

This study had several limitations. First, the homogeneity of the study sample may hinder generalization of the results to functional recovery outcomes in patients with advanced cancer who already had impaired functioning presurgery (ie, an elevated baseline interference score driven by disease plus previous cancer therapy). However, the methods established herein may be applicable to patients undergoing curative surgery for other types of early-stage cancer or to patients who have good presurgery functional status, as is more likely to be found in noncancer populations. Second, an individual’s interpretation of “general activity” in the past 24 hours could vary―for example, it could encompass bathing, self-care, driving, or taking care of family―and this could explain the item’s insensitivity to detect time to return to baseline by type of procedure. However, this item was highly sensitive to dynamic change in the severity of functional interference over time, providing a robust postoperative physical-activity measure that could be more generally applicable than a more-specific PRO measure would be (for example, some may not need to drive, some may not need to take care of a family). Third, we did not collect data on the extent of resection, which could have affected patient recovery. However, no significant between-group differences were found in terms of preoperative comorbidity, cancer stage, or performance status, suggesting the comparability of the open surgery and VATS samples. Finally, the study lacked an objective measure of functional impairment by which to define recovery to “good” functional status (both cutpoints and minimal clinically important difference), other than by charting the return to presurgery PRO levels (in this sample with early-stage NSCLC, most patients rated their baseline interference as 0). Compared with the long time needed to return to unimpaired presurgery functional levels, the option for using the time to return to “good” functioning status could be more practical for further decision making about ongoing cancer care (such as when to start chemotherapy), and in that case there would be no need for a baseline assessment.

In conclusion, the current study is among the first to establish PROs for describing the nature of 2 domains of functional recovery, physical and psychological, in patients with early-stage NSCLC during the postoperative period. As a simple, concise PRO assessment instrument, the MDASI is a clinically relevant and user-friendly tool for obtaining the patient’s perspective on how well he or she is recovering from surgery and for optimizing perioperative care (13). Routine monitoring of functional recovery on the basis of MDASI interference scores, especially the “walking” item, would be a novel PRO application in perioperative care that has the potential to improve standard practice once patients have been discharged after major cancer surgery.