Impact of the Extent of Thyroidectomy on Quality of Life in Differentiated Thyroid Cancer Survivors: A Propensity Score Matched Analysis

Figures & data

Table 1 Domain Division and Domain-Based Scoring Method for EORTC QLQ-C30

Domain	Property	Item Numbers	R	RS	SS
Physical functioning (PF)	Functioning	5	3	$\frac{Q_1+Q_2+Q_3+Q_4+Q_5}{5}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Emotional functioning (EF)	Functioning	4	3	$\frac{Q_{21}+Q_{22}+Q_{23}+Q_{24}}{4}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Cognitive functioning (CF)	Functioning	2	3	$\frac{Q_{20}+Q_{25}}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Social functioning (SF)	Functioning	2	3	$\frac{Q_{26}+Q_{27}}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Role functioning (RF)	Functioning	2	3	$\frac{Q_6+Q_7}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Fatigue (FA)	Symptom	3	3	$\frac{Q_{10}+Q_{12}+Q_{18}}{3}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Nausea & vomiting (NV)	Symptom	2	3	$\frac{Q_{14}+Q_{15}}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Pain (PA)	Symptom	2	3	$\frac{Q_9+Q_{19}}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Dyspnoea (DY)	Symptom	1	3	$\text{{Q\_8}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Diarrhoea (DI)	Symptom	1	3	$\text{{Q\_{17}}}$	$\left(-\frac{\text{RS}-1}{R}\right)*100$
Constipation (CO)	Symptom	1	3	$\text{{Q\_{16}}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Financial difficulties (FD)	Symptom	1	3	$\text{{Q\_{28}}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Insomnia (IN)	Symptom	1	3	$\text{{Q\_{11}}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Lack of appetite (LA)	Symptom	1	3	$\text{{Q\_{13}}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Global quality of life (GQ)	Overall	2	6	$\frac{Q_{29}+Q_{30}}{2}$	$\left(\frac{\text{RS}-1}{R}\right)*100$

Note: R=maximum score-minimum score.

Table 2 Domain Division and Domain-Based Scoring Method for THYCA-QoL

Domain	Item Numbers	R	RS	SS
Neuromuscular (NM)	3	3	$\frac{Q_{10}+Q_{12}+Q_{13}}{3}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Voice (VO)	2	3	$\frac{Q_3+Q_4}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Concentration (CT)	2	3	$\frac{Q_{20}+Q_{21}}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Sympathetic (ST)	2	3	$\frac{Q_8+Q_9}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Throat/mouth problems (TM)	3	3	$\frac{Q_1+Q_2+Q_5}{3}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Psychological (PC)	4	3	$\frac{Q_{17}+Q_{19}+Q_{22}+Q_{23}}{4}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Sensory (SE)	2	3	$\frac{Q_{15}+Q_{16}}{2}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Scar (SC)	1	3	$\text{{Q\_6}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Chilly (CH)	1	3	$\text{{Q\_7}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Tingling hands/feet (THF)	1	3	$\text{{Q\_{11}}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Gained weight (GW)	1	3	$\text{{Q\_{14}}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Headache (HA)	1	3	$\text{{Q\_{18}}}$	$\left(1-\frac{\text{RS}-1}{R}\right)*100$
Interest of sex (SXI)	1	3	$\text{{Q\_{24}}}$	$\left(\frac{\text{RS}-1}{R}\right)*100$

Note: R=maximum score-minimum score.

Figure 1 QoL bifactor model for the EORTC QLQ-C30.

Figure 2 QoL bifactor model for the THYCA-QoL.

Table 3 Fit Indices for the Bifactor QoL Model

Model	df	RMSEA	RMSEA_5	RMSEA_95	SRMSR	TLI	CFI
EORTC QLQ-C30
Single-factor model	351	0.0869	0.0780	0.0952	0.1058	0.7976	0.8135
Three-factor model	351	0.0903	0.0815	0.0985	0.2062	0.7812	0.7985
Bifactor model	321	0.0559	0.0445	0.0662	0.0913	0.9163	0.9295
THYCA QoL
Single-factor model	206	0.0902	0.0788	0.1011	0.0982	0.5854	0.6287
Three-factor model	206	0.0949	0.0837	0.1056	0.1834	0.5412	0.5890
Bifactor model	182	0.0568	0.0417	0.0704	0.0772	0.8358	0.8700

Table 4 Participant Characteristics Before and After the Propensity Score Matching

	Before PS Matching				After PS Matching
	Untreated No.(%)	Treated No.(%)	Total N	p value	Untreated No.(%)	Treated No.(%)	Total N	p value
Age/years	40 (10.2, 23, 67)	41 (10.2, 24, 67)	150	0.844	41 (9.42, 27, 67)	40 (8.5, 24, 61)	133	0.821
Income level
≤3000	19 (20.9)	10 (16.9)	29	0.446	17 (21.5)	9 (16.7)	26	0.521
3000~5000	17 (18.7)	11 (18.6)	28		15 (19)	11 (20)	26
≥5000	55 (60.4)	38 (64.4)	93		47 (59.5)	34 (63)	81
Employed status
No	36 (39.6)	26 (44.1)	62	0.700	33 (42)	21 (39)	54	0.915
Yes	55 (60.4)	33 (55.9)	88		46 (58)	33 (61)	79
Marital status
No	12 (13.2)	1 (1.7)	13	0.046	0 (0)	0 (0)	0	–
Yes	79 (86.8)	58 (98.3)	137		79 (100)	54 (100)	133
Means of payment
Own expense	4 (4.4)	5 (8.5)	9	0.256	4 (5.1)	4 (7.4)	8	0.231
Medical Insurance (Urban)	67 (73.6)	41 (69.5)	108		57 (72)	38 (70)	95
Medical Insurance (Rural)	4 (4.4)	7 (11.9)	11		3 (3.8)	7 (13)	10
Business Insurance	0 (0)	1 (1.7)	1		0 (0)	0 (0)	0
Other	16 (17.6)	5 (8.5)	21		15 (19)	5 (9.3)	20
Total	91	59	150		79	54	133

Note: Continuous data are presented as mean (standard deviations, minimum, maximum), and categorical data are presented as frequencies with percentages.

Figure 3 Common support region of propensity scores.

Table 5 Covariate Balance Before and After Propensity Score Matching

Covariates	Before PS Matching			After PS Matching
	Mean in Treated Group	Mean in Untreated Group	SMD(%)	Mean in Treated Group	Mean in Untreated Group	SMD(%)
Age/years	41.407	39.846	15.3	40.463	40.958	−4.9
Income	2.4746	2.3956	9.9	2.463	2.4523	1.3
Employed	0.5593	0.6044	−9.1	0.611	0.5664	9.0
Marital	0.9831	0.8681	44.6	1	1	0.0
Payment	2.322	2.5275	−18.9	2.333	2.358	−2.3

Table 6 Summary of the Distribution of the Bias

Sample	Pseudo R^2	LR X^2	$\text{Prob gt}{\chi }^2$	MeanBias (SMD)	MedBias
Unmatched	0.045	9.11	0.105	19.6	15.3
Matched	0.002	0.27	0.992	3.5	2.3

Figure 4 Standard mean deviation of covariates.

Table 7 Impact of Surgery Type on the QoL in DTC Patients Before the Propensity Matched Analysis

		EORTC QLQ-C30			THYCA-QoL
		Coef.	Std. Err.	$\text{p}$ value	Coef.	Std. Err.	$\text{p}$ value
Sum scoring	Univariate analysis	−0.448	1.506	0.776	−0.416	1.226	0.734
	Multivariate analysis	−0.904	1.587	0.569	−0.272	1.267	0.830
Domain-based scoring	Univariate analysis	−0.824	1.575	0.602	−1.083	1.693	0.523
	Multivariate analysis	−1.135	1.676	0.499	−0.841	1.752	0.632
IRT-based scoring	Univariate analysis	−0.401	0.225	0.076	0.044	0.233	0.852
	Multivariate analysis	−0.512	0.229	0.027	0.091	0.247	0.713

Table 8 Impact of Surgery Type on the QoL in DTC Patients After the Propensity Matched Analysis

		EORTC QLQ-C30			THYCA-QoL
		Coef.	Std. Err.	$\text{p}$ value	Coef.	Std. Err.	$\text{p}$ value
Sum scoring	PSM analysis (ATE)	−0.400	1.906	0.834	−0.827	1.516	0.586
	PSM analysis (ATT)	−0.655	2.191	0.765	−0.817	1.739	0.639
	PSM analysis (ATU)	−0.226	2.107	0.914	−0.834	1.643	0.612
Domain-based scoring	PSM analysis (ATE)	−0.491	2.015	0.807	−1.692	2.036	0.406
	PSM analysis (ATT)	−0.703	2.358	0.766	−1.788	2.369	0.450
	PSM analysis (ATU)	−0.347	2.193	0.874	−1.627	2.312	0.482
IRT-based scoring	PSM analysis (ATE)	−0.442	0.279	0.114	−0.032	0.278	0.908
	PSM analysis (ATT)	−0.480	0.308	0.119	−0.062	0.319	0.845
	PSM analysis (ATU)	−0.416	0.306	0.174	−0.011	0.307	0.971

Table 9 Impact of Surgery Type on the Domain of EORTC QLQ-C30 in DTC Patients After the Propensity Matched Analysis

	Domain	Coef.	Std. Err.	$\text{p}$ value	Domain	Coef.	Std. Err.	$\text{p}$ value
PSM analysis (ATE)	FA	3.613	3.838	0.346	LA	−0.351	1.268	0.782
PSM analysis (ATT)		3.873	4.293	0.367		−0.079	1.368	0.954
PSM analysis (ATU)		3.436	4.186	0.412		−0.538	1.436	0.708
PSM analysis (ATE)	NV	−0.500	2.046	0.807	PF	−0.118	1.725	0.946
PSM analysis (ATT)		−0.518	2.539	0.939		−0.147	1.917	0.939
PSM analysis (ATU)		−0.488	2.116	0.817		−0.098	1.909	0.959
PSM analysis (ATE)	PA	1.450	3.167	0.647	EF	2.821	3.238	0.384
PSM analysis (ATT)		1.885	3.741	0.614		2.556	3.593	0.477
PSM analysis (ATU)		1.154	3.519	0.743		3.001	3.634	0.409
PSM analysis (ATE)	DY	3.849	4.169	0.356	CF	0.962	3.712	0.796
PSM analysis (ATT)		3.462	4.669	0.458		0.599	4.466	0.893
PSM analysis (ATU)		4.114	4.684	0.38		1.21	3.882	0.755
PSM analysis (ATE)	DI	−1.027	3.289	0.755	SF	−1.384	3.73	0.711
PSM analysis (ATT)		−1.519	3.494	0.664		−3.361	4.193	0.423
PSM analysis (ATU)		−0.691	3.685	0.851		−0.033	4.06	0.994
PSM analysis (ATE)	CO	−4.420	3.889	0.256	RF	−4.621	3.231	0.153
PSM analysis (ATT)		−3.193	4.554	0.483		−4.433	3.523	0.208
PSM analysis (ATU)		−5.259	4.136	0.204		−4.750	3.593	0.186
PSM analysis (ATE)	FD	1.318	2.929	0.653	GQ	−1.245	3.939	0.95
PSM analysis (ATT)		1.313	3.223	0.684		−0.681	4.184	0.871
PSM analysis (ATU)		1.321	3.34	0.693		0.053	4.565	0.991
PSM analysis (ATE)	IN	−0.054	2.703	0.984
PSM analysis (ATT)		0.687	2.957	0.816
PSM analysis (ATU)		0.379	3.108	0.9903

Table 10 Impact of Surgery Type on the Domain of THYCA-QoL in DTC Patients After the Propensity Matched Analysis

	Domain	Coef.	Std. Err.	$\text{p}$ value	Domain	Coef.	Std. Err.	$\text{p}$ value
PSM analysis (ATE)	NM	1.67	2.912	0.566	SC	7.536	4.97	0.129
PSM analysis (ATT)		0.874	3.174	0.783		8.5	5.288	0.108
PSM analysis (ATU)		2.214	3.295	0.502		6.876	5.69	0.227
PSM analysis (ATE)	VO	6.078	4.444	0.171	CH	3.062	5.33	0.566
PSM analysis (ATT)		6.013	5.085	0.237		4.511	5.563	0.417
PSM analysis (ATU)		6.123	4.741	0.197		2.071	6.42	0.747
PSM analysis (ATE)	CT	−1.654	3.936	0.674	THF	4.928	4.228	0.244
PSM analysis (ATT)		−2.341	4.802	0.626		5.359	4.778	0.262
PSM analysis (ATU)		−1.185	4.098	0.772		4.633	4.715	0.326
PSM analysis (ATE)	ST	1.478	4.133	0.721	GW	−2.366	5.574	0.671
PSM analysis (ATT)		1.254	4.34	0.778		−2.491	6.106	0.683
PSM analysis (ATU)		1.632	4.626	0.778		−2.280	6.308	0.718
PSM analysis (ATE)	TM	−2.002	3.674	0.586	HA	1.217	4.181	0.771
PSM analysis (ATT)		−1.779	4.141	0.667		0.977	4.836	0.84
PSM analysis (ATU)		−2.154	3.993	0.59		1.381	4.414	0.754
PSM analysis (ATE)	PC	0.545	2.967	0.854	SXI	−3.377	4.168	0.418
PSM analysis (ATT)		0.551	3.214	0.864		−3.558	4.608	0.44
PSM analysis (ATU)		0.54	3.383	0.873		−3.253	4.53	0.473
PSM analysis (ATE)	SE	−2.425	2.866	0.397
PSM analysis (ATT)		−2.689	3.449	0.436
PSM analysis (ATU)		−2.245	3.038	0.46