Overuse in surgery by deviation from the “Choosing Wisely” campaign recommendations: a surrogate of unnecessary surgery and low-value care

Abstract

Background

Inappropriate health care reduces the quality of care and efficiency of the health system. The goal of this study was to analyze the frequency of overuse of five surgical procedures while also determining the epidemiological characteristics of both patients and interventions.

Methods

This is a retrospective observational cohort study performed in a tertiary referral center in Spain. Surgical Overuse was analyzed according to five “Choosing Wisely” recommendations related to both interventions and perioperative care. The association between overuse and the characteristics of patients and procedures was assessed by bivariate and multivariate predictive analysis. Costs were calculated for each clinical episode using data from the hospital’s accounting department; total hospitalization costs from the index procedure to hospital discharge were used for intervention-related overuse, and average unit cost for perioperative practices.

Results

A total of 895 clinical episodes were analyzed over one year. Low-value surgery was identified in 9.7% (N: 87; 95% CI: 7.9 to 11.9). Specifically, overuse was identified in 59.6% (53 out of 89; 48.6 to 69.7)) of the echocardiograms performed after valve replacement, 15.7% (26 out of 166; 10.6 to 22.3) of podiatric surgeries, and 2.0% (8 out of 400; 0.9 to 4.1) of cholecystectomies for asymptomatic cholelithiasis. No overuse was found among Mohs surgeries or in the perioperative use of opioids for pediatric patients. The risk of overuse was higher in patients with the following characteristics: increased age (OR [95% CI]: 1.1 [1.1 to 1.2] for each additional year); ≥1 intrinsic risk factor (IRF) (3.5 [1.5 to 8.0] versus absence of IRF); and hospitalized patients (12.8 [4.5 to 37.1] versus ambulatory surgery). Overuse represented an overall cost of €86,858.05.

Conclusion

Applying five “Choosing Wisely” recommendations, low-value surgical practices were identified, mainly involving echocardiograms after valve replacement, podiatric surgery, and cholecystectomies. Such practices were more frequent among older and hospitalized patients and were associated with direct economic costs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13037-026-00477-z.

Background

Health inappropriateness is defined as the adoption of practices whose risks outweigh the benefits, are inefficient, or are not adapted to patients’ needs [1]. It can manifest in the form of underuse, when a medical service that would have produced a favourable result for the patient is not provided [2], or overuse, when a non-cost-effective procedure is performed [3].

Overuse has been identified in practically all health systems, manifesting in a variety of forms, such as unnecessary diagnostic and screening testing or imaging studies, inappropriate surgical procedures or overprescription of medications [3]. It adds no potential value to medical processes and patients; instead, it can be potentially harmful, consume economic and operating room resources, and raise ethical concerns within a system with limited resources [4, 5]. Surgical overuse has been shown to be a widespread phenomenon associated with high additional expenses. For example, it has been identified in 10,4% of percutaneous coronary angiograms performed in patients with stable coronary artery disease in the United States [6], in 73.0% of preoperative chest radiographs in Spain [7], and 20–40% of interventions related to pancreatic cancer in Italy [8].

In recent years, prestigious institutions, in collaboration with various scientific societies, have promoted numerous initiatives to improve the quality of care through evidence-based recommendations. “Choosing Wisely”, launched in 2012 in the United States by the American Board of Internal Medicine (ABIM) Foundation, stands out as an international initiative aimed at enhancing clinical decision-making and reducing inappropriate practices [9–11].

Currently, few studies explore overuse in surgery, and those are usually limited to the analysis of a single procedure in a very specific healthcare context [8, 12]. Furthermore, their traditional approach predominantly focuses on the study of care-related outcome measures without considering potential risk factors related to the patient or the intervention.

Therefore, this study aims to analyze the frequency of overuse in five surgical procedures (hallux valgus/hammer toe interventions, cholecystectomies, Mohs surgeries, echocardiograms after valve replacement and use of opioids during the perioperative management of pediatric patients.) by applying “Choosing Wisely” recommendations. The goal is to expand our knowledge about the effect of overuse on the patient and its associated economic impact.

Methods

A single-center, retrospective observational cohort study was performed to examine the frequency, determinants and costs of low-value care surgery. We hypothesized that overuse in surgery is frequent and associated with identifiable patient and procedural characteristics.

This study was conducted at Ramón y Cajal University Hospital, a tertiary referral center located in the Community of Madrid (Madrid, Spain), equipped with 44 operating rooms and 795 licensed beds.

All “Choosing Wisely” recommendations developed by surgical scientific societies that were available on their website as of 12/12/2020 [9] were compiled. Recommendations explicitly referring to the performance or perioperative management of surgical interventions were selected.

Afterwards, four opportunistic criteria were applied for the final selection of recommendations: (1) the recommendations represented various surgical specialities; (2) the recommendations targeted various age groups; (3) the recommendations addressed different perioperative phases; (4) the quality of the data in the associated clinical episodes for each recommendation was deemed sufficient.

Finally, five recommendations were selected. For three recommendations, the overuse identified was related to the performance of a surgical intervention.

1. Don’t undergo surgery for a bunion or hammertoes without symptoms [13].

2. Avoid routine cholecystectomy for patients with asymptomatic cholelithiasis [14].

3. Do not treat uncomplicated, nonmelanoma skin cancer less than 1 cm in size on the trunk and extremities with Mohs micrographic surgery [15].

For the other two recommendations, the overuse involved the performance of a practice related to the perioperative clinical management of the patient:

• 4.Do not perform a routine predischarge echocardiogram after cardiac valve replacement surgery [16].
• 5.Reduce post-operative opioid requirements in pediatric patients by administering acetaminophen and/or non-steroidal anti-inflammatory medications in the perioperative period [17].

Afterwards, clinical episodes were eligible for inclusion if they involved one of the interventions specified in the selected recommendations and occurring between July 1, 2018 and June 30, 2019.

All clinical episodes were initially identified through Diagnosis-Related Group (DRG) codes [18] by the Information Systems Unit of the Admission Coordination, Care and Clinical Documentation Information Systems.

The following inclusion criteria were applied: (1) clinical episodes with codes corresponding to surgical interventions for hallux valgus and hammertoe; (2) clinical episodes with codes corresponding to cholecystectomy; (3) clinical episodes involving dermatological surgical procedures; (4) clinical episodes with codes corresponding to cardiac valve replacement; and (5) clinical episodes involving surgical procedures in pediatric patients.

Subsequently, the medical records of all episodes associated to the selected recommendations were reviewed. Episodes in which the intervention of interest had not actually been performed were excluded.

The final study sample consisted of all clinical episodes with associated procedures that met the predefined selection criteria and, after chart review, were confirmed to include the intervention of interest. (Fig. 1).

Fig. 1.

Diagram of the selection process for recommendations and sample. (%): Frequency of procedure overuse. (1) “Choosing Wisely” recommendation applied to determine overuse: Don’t undergo surgery for a bunion or hammertoes without symptoms. (2) “Choosing Wisely” recommendation applied to determine overuse: Avoid routine cholecystectomy for patients with asymptomatic cholelithiasis. (3) “Choosing Wisely” recommendation applied to determine overuse: Do not treat uncomplicated, nonmelanoma skin cancer less than 1 cm in size on the trunk and extremities with Mohs micrographic surgery. (4) “Choosing Wisely” recommendation applied to determine overuse: Do not perform a routine predischarge echocardiogram after cardiac valve replacement surgery. (5) “Choosing Wisely” recommendation applied to determine overuse: Reduce post-operative opioid requirements in paediatric patients by administering acetaminophen and/or non-steroidal anti-inflammatory medications in the perioperative period

The following outcome measures were collected (Fig. 2, Supplementary Table 1 [see Additional file 1]):

Fig. 2.

Outcome measures according to the healthcare phase

At the time of hospital admission: sex, age, degree of surgical risk according to the American Society of Anesthesiologists (ASA) Physical Status scale [19] and the Charlson index [20]. The presence of intrinsic risk factors (IRFs) associated with the development of complications and adverse events was documented based on criteria established by highly regarded studies in the field of patient safety [21, 22]. These factors show patients’ baseline vulnerability and clinical complexity and help to contextualize surgical outcomes. The following variables were included: kidney failure, diabetes mellitus, cardiovascular disease, chronic lung disease, immunodeficiency, active smoking, mobility disorders and sensory deficits, neutropenic neoplasia, cirrhosis, hypoalbuminemia, pressure ulcers, and obesity (The definitions of these IRFs are found in the Supplementary Table 1 [see Additional file 1]). Additionally, the final discharge destination of each patient was recorded.

Related to the intervention: administrative type, origin, intervention shift, and degree of contamination of the surgery.

Related to overuse: the existence of low-value care according to the selected recommendations; costs derived from such overuse.

The primary outcome considered of this study was the frequency of low-value surgical procedures. Secondary outcomes included the association between overuse in surgery and patient-related and intervention-related characteristics, as well as the economic impact of low-value surgical care.

Doctors with specific training conducted all clinical episode review work.

Statistical analysis

A descriptive and association analysis of overuse in relation to the outcome measures of interest was performed, as follows: (1) at the overall sample; (2) stratifying by each procedure.

Proportions and 95% confidence intervals (95% CI) were estimated for qualitative variables. For quantitative variables, the median and interquartile range (IQR) were reported, as the normality assumption was not fulfilled according to the Kolmogorov‒Smirnov test. Bivariate analysis was performed on qualitative parametric variables via the χ2 test and on nonparametric variables via Fisher’s exact test. Statistical significance was established at p < 0.05 for all tests. Twelve records were excluded from the bivariable analysis of the ASA grade due to missing data (9 from Mohs surgeries and 3 from pediatric interventions).

A multivariate predictive analysis using logistic regression was conducted to analyze overuse, adjusting for the variables of interest. A backward modelling strategy was applied, with a statistical significance threshold set at p < 0.1. Resampling techniques (Bootstrap) were employed to address potential overoptimism. The goodness of fit was assessed using the Hosmer-Lemeshow test. The global multivariate analysis did not account for the specific type of procedure performed. However, additional models were developed for procedures where overuse was identified (for podiatric surgery, cholecystectomies and echocardiograms after valve replacement).

For the recommendations where overuse referred to the inappropriate performance of a surgical intervention, the degree of overuse was calculated as the percentage of these interventions performed unnecessarily over the total number of interventions of the same type performed during the entire study period.

For the recommendations where overuse referred to a specific perioperative management for a particular type of intervention, the degree of overuse was calculated as the percentage of these interventions in which such unjustified perioperative management was performed.

All statistical analyses were performed in R version 4.0.5 [23].

Cost analysis

The costs associated with overuse were calculated (1) for the overall sample; (2) stratified by whether the unjustified procedure was an intervention or part of the perioperative management; (3) for each of the five procedures.

For clinical episodes associated to recommendations related to interventions (podiatric surgery, cholecystectomy and Mohs surgery), total costs attributable to overuse were calculated based on hospitalization-related expenses, such as inpatient stay, personnel, facilities, medical supplies, and diagnostic or therapeutic procedures. These costs were calculated from the time of the index procedure until hospital discharge from data provided by the hospital’s accounting department.

For clinical episodes associated to recommendations related to perioperative management (echocardiograms after valve replacement and the use of opioids in pediatric patients), only the cost of the inappropriate procedure itself was considered. It was calculated using its average unit cost from data provided by the hospital’s accounting department.

Results

Sample characteristics

A final sample of 895 episodes that contained an intervention described by one of the five selected recommendations was obtained.

A total of 56.4% (N: 505) of the patients were women, the median age was 68 (IQR: 55–75) years, and 75.5% (667) had an ASA I-II grade. In addition, 65.1% (583) of the patients had ≥ 1 IFR. A total of 78.7% (704) of the interventions were performed for patients on the waiting list, and 55.5% (497) had a scheduled hospitalization (Tables 1 and 2).

Table 1.

Degree of overuse according to patient characteristics and destination at discharge

	Overall		Overuse
	n (%)	n	%	95% CI	p
Sex					0.6
Female	505 (56.4%)	47	9.3%	7.0 to 12.3
Male	390 (43.6%)	40	10.3%	7.5 to 13.8
Categorical age					< 0.001***
≤ 18 years	135 (15.1%)	0	0.0%	0.0 to 3.4
19 to 50 years	156 (17.4%)	12	7.7%	4.2 to 13.0
51 to 65 years	224 (25.0%)	26	11.6%	7.9 to 16.7
66 to 0 years	286 (32.0%)	42	14.7%	10.9 to 19.4
> 80 years	94 (10.5%)	7	7.4%	3.3 to 15.2
ASA grade (1)					< 0.001***
I to II	667 (75.5%)	32	4.8%	3.4 to 6.8
III to IV	216 (24.5%)	55	25.5%	19.9 to 31.9
Intrinsic risk factors					< 0.001***
0	312 (34.9%)	10	3.2%	1.6 to 6.0
≥ 1	583 (65.1%)	77	13.2%	10.6 to 16.3
Charlson index					< 0.001***
0	406 (45.4%)	22	5.4%	3.5 to 8.2
1	190 (21.2%)	26	13.7%	9.3 to 19.6
2 to 3	237 (26.5%)	34	14.3%	10.3 to 19.6
> 3	62 (6.9%)	5	8.1%	3.0 to 18.5
Destination upon discharge					0.002**
Speciality center	584 (65.3%)	67	11.5%	9.1 to 14.4
Primary care	301 (33.6%)	18	6.0%	3.7 to 9.5
Other service	6 (0.7%)	0	0.0%	0.0 to 48.3
Death	2 (0.2%)	0	0.0%	0.0 to 80.2
Other hospital	2 (0.2%)	2	100%	19.8 to 100
Total	895 (100%)	87	9.7%	7.9 to 11.9

n: sample, 95% CI: 95% confidence interval. P-value: Obtained from the chi-square test (parametric) or Fisher's exact test (nonparametric).*p<0.05; **p<0.01;***p<0.001. (1) ASA: 12 records with missing data were excluded from the analysis

Table 2.

Degree of overuse according to the characteristics of the intervention

	Overall		Overuse
	n (%)	n	%	95% CI	p
Procedure					< 0.001***
Cholecystectomies(1)	400 (44.7%)	8	2.0%	0.9 to 4.1
Hallux valgus/hammer toe surgery(2)	166 (18.5%)	26	15.7%	10.6 to 22.3
Use of perioperative opioids in pediatric patients(3)	130 (14.5%)	0	0.0%	0.0 to 3.6
Mohs surgery(4)	110 (12.3%)	0	0.0%	0.0 to 4.2
Echocardiograms after valve replacement(5)	89 (9.9%)	53	59.6%	48.6 to 69.7
Intervention type					< 0.001***
Scheduled with hospitalization	497 (55.5%)	63	12.7%	9.9 to 16.0
Scheduled without hospitalization	354 (39.6%)	17	4.8%	2.9 to 7.7
Urgent	44 (4.9%)	7	15.9%	7.2 to 30.7
Intervention origin					< 0.001***
Waiting list	704 (78.7%)	63	8.9%	7.0 to 11.4
Minor unscheduled surgery	122 (13.6%)	6	4.9%	2.0 to 10.8
Hospitalization	62 (6.9%)	18	29.0%	18.6 to 42.1
Emergency	7 (0.8%)	0	0.0%	0.0 to 43.9
Shift of the intervention					0.022*
Morning	647 (72.3%)	72	11.1%	8.9 to 13.9
Afternoon/Night	248 (27.7%)	15	6.0%	3.5 to 10.0
Contamination					0.002**
Clean	475 (53.1%)	60	12.6%	9.8 to 16.0
Clean-contaminated	383 (42.8%)	27	7.0%	4.8 to 10.2
Contaminated	37 (4.1%)	0	0.0%	0.0 to 11.7
Total	895 (100%)	87	9.7%	7.9 to 11.9

n: sample, 95% CI: 95% confidence intervals. P-value: Obtained from the chi-square test (parametric) or Fisher's exact test (nonparametric).*p<0.05; **p<0.01;***p<0.001. (1) “Choosing Wisely” recommendation applied to determine overuse: Avoid routine cholecystectomy for patients with asymptomatic cholelithiasis. (2) “Choosing Wisely” recommendation applied to determine overuse: Don't undergo surgery for a bunion or hammertoes without symptoms. (3) “Choosing Wisely” recommendation applied to determine overuse: Reduce post-operative opioid requirements in pediatric patients by administering acetaminophen and/or non-steroidal anti-inflammatory medications in the perioperative period(4) “Choosing Wisely” recommendation applied to determine overuse: Do not treat uncomplicated, nonmelanoma skin cancer less than 1 cm in size on the trunk and extremities with Mohs micrographic surgery.(5) “Choosing Wisely” recommendation applied to determine overuse: Do not perform a routine predischarge echocardiogram after cardiac valve replacement surgery

A description of the sample for each procedure is included in Supplementary Table 2 [see Additional file 1].

Overuse characteristics

Overall, overuse was observed in 9.7% (87 out of 895; 95% CI: 7.9 to 11.9) of the episodes analyzed. Specifically, overuse was observed in 59.6% (53 out of 89; 48.6 to 69.7) of the episodes involving the performance of echocardiograms after valve replacement, followed by 15.7% (26 out of 166; 10.6 to 22.3) involving the podiatric surgery and 2.0% (8 out of 400; 0.9 to 4.1) involving cholecystectomy. No evidence of overuse was found for Mohs surgery or the perioperative use of opioids in pediatric patients (Table 2).

In the bivariable analysis, overuse was higher among patients with ASA grade III-V (25.5% versus 4.8% among ASA grade I-II; p < 0.001) or with ≥ 1 IRF (13.2% versus 3.2% among 0 IRFs; p < 0.001). A higher degree of overuse was also observed in patients referred to another hospital (100%) or in those whose follow-up was conducted in specialized care (11.5%) compared to patients who required primary care follow-up (6.0%; p = 0.002) (Table 1).

Overuse was higher in interventions performed on already hospitalized patients (29.0% versus 8.9% among those from the waiting list, and 4.9% among minor unscheduled surgeries; p < 0.001), as well as in scheduled interventions with hospitalization (12.7% versus 4.8% among those scheduled without hospitalization; p < 0.001) (Table 2 and Supplementary Table 3 [see Additional file 1]).

Multivariate analysis

For the overall sample, after adjusting for sex, age, IRFs, intervention origin and shift, the risk of overuse increased by 1.1 times for each additional year of patient age (OR [95% CI]: 1.1 [1.1 to 1.2], p = 0.01) and by 3.5 times for patients with ≥ 1 IRF (compared to 0 IRFs; 3.5 [1.5 to 8.0] p = 0.002). Furthermore, the risk of overuse increased by 4.5 times in patients on the waiting list (4.5 [1.5 to 913.5]; p = 0.007) and by 12.8 times in hospitalized patients (12.8 [4.5 to 37.1]; p < 0.001) compared to minor unscheduled surgeries. A Hosmer-Lemeshow test p-value of 0.24 was obtained.

For interventions involving podiatric surgery, the risk of overuse was 6.1 times greater in patients with neoplasia (6.1 [1.2 to 27.6]; p = 0.020) and 14.0 times higher in those with pulmonary disease (14.0 [3.4 to 58.2]; p < 0.001) compared to patients without these IRFs; the p-value of the Hosmer–Lemeshow test was 0.06 (Table 3). No significant model was obtained for the independent analysis of overuse in cholecystectomies and echocardiograms after valve replacement.

Table 3.

Risk of overuse from regression models adjusted for patient and intervention characteristics and by procedure

Overall(1)	n (%)	OR	95% CI	p
Age
Increase of 1 year, mean in years (S)	67.0 (12.0)	1.1	1.1 to 1.2	0.01*
Intrinsic risk factors
0	10 (11.5%)	1.0	NA	NA
≥ 1	77 (88.5%)	3.5	1.5 to 8.0	0.002**
Intervention origin
Minor unscheduled surgery	6 (6.9%)	1.0	NA	NA
Waiting list	63 (72.40%)	4.5	1.5 to 13.5	0.006**
Hospitalization	18 (20.7%)	12.8	4.5 to 37.1	< 0.001***
Emergency	0 (0.0%)	NA	NA	NA
Intervention shift
Morning	72 (82.8%)	1.0	NA	NA
Afternoon/Night	15 (17.2%)	0.3	0.2 to 0.7	0.003**
Constant		0.0	0.0 to 0.02	< 0.001***
Hallux valgus/hammer toe surgery (2)	n (%)	OR	95% CI	p
Neoplasia (3)	2 (40%)	6.1	1.2 to 27.6	0.020*
Lung disease (3)	8 (47.1%)	14.0	3.4 to 58.2	< 0.001***
Charlson index
0	12 (15.0%)	1.0	NA	NA
1	6 (14.3%)	0.3	0.1 to 0.9	0.036*
2–3	8 (18.2%)	0.6	0.2 to 2.0	0.387
Constant		0.2	0.1 to 0.3	< 0.001***

OR = Odds Ratio, 95 CI: 95% Confidence Interval; p: estimated from the logistic regression. S: standard deviation(1) Multivariate logistic regression analysis of overuse, adjusting for age, intrinsic risk factors, and the origin and shift of the intervention for the overall sample.(2) Multivariate logistic regression analysis of overuse, adjusting for intrinsic risk factors in hallux valgus/hammer toe intervention. “Choosing Wisely” recommendation applied to determine overuse: Don't undergo surgery for a bunion or hammertoes without symptoms.(3) Reference value: absence of the intrinsic risk factors.*p<0.1; **p<0.01;***p<0.001

Economic impact

The overall cost associated with overuse was €86,858.05. For each of the analyzed procedures, the total expenses were as follows: €58,279.93 for 26 podiatric surgeries (mean cost: €2,241.54 per low-value procedure); €26,304.26 for 8 cholecystectomies (mean cost: €3,288.03 per per low-value procedure); and €2,273.86 for 53 echocardiograms after valve replacement (mean cost: €42.90 per low-value procedure) (Table 4).

Table 4.

Costs of overuse

Procedure type		Procedure	Overuse (n)	Mean cost	Total
Intervention	1	Surgery on hallux valgus and hammer toes(1)	26	2241.54€	58279.93€	84584.19€	86858.05€
	2	Cholecystectomies(2)	8	3288.03€	26304.26€
	3	Mohs Surgery(3)	-	-	-
Perioperative care	4	Echocardiograms after valve replacement interventions(4)	53	42.90€	2273.86€	2273.86€
	5	Use of perioperative opioids in pediatric patients(5)	-	-	-

(1) “Choosing Wisely” recommendation applied to determine overuse:  Don't undergo surgery for a bunion or hammertoes without symptoms.(2) “Choosing Wisely” recommendation applied to determine overuse:  Avoid routine cholecystectomy for patients with asymptomatic cholelithiasis.(3) “Choosing Wisely” recommendation applied to determine overuse: Do not treat uncomplicated, nonmelanoma skin cancer less than 1 cm in size on the trunk and extremities with Mohs micrographic surgery.(4) “Choosing Wisely” recommendation applied to determine overuse:  Do not perform a routine predischarge echocardiogram after cardiac valve replacement surgery(5) “Choosing Wisely” recommendation applied to determine overuse: Reduce post-operative opioid requirements in paediatric patients by administering acetaminophen and/or non-steroidal anti-inflammatory medications in the perioperative period

Discussion

Unjustified procedures were identified in approximately 10% of the clinical episodes. The highest rates of overuse were observed in recommendations related to echocardiograms after valve replacement (59.6%), podiatric surgery (15.7%) and cholecystectomies (2.0%). Conversely, no overuse was detected in relation to Mohs surgery or the perioperative use of opioids in pediatric patients. Overuse was greater among older patients, those with ≥ 1 IRF, and interventions performed on hospitalized patients and conducted during the morning shift. The total cost associated with overuse was 86,858.05 euros.

The overall degree of overuse (9.7%) is consistent with findings from a series of studies performed in the United States and Canada between 2011 and 2015 which analyzed 28 recommendations not limited to surgical procedures, as well as with a large Canadian meta-analysis published in 2022 (4.0% to 14.9%) [24–27]. However, these results are not usually static, as the degree of inappropriateness may fluctuate over time, as demonstrated in an analysis conducted in Australia between 2010 and 2014 on 21 hospital procedures [28].

More specifically, the high degree of overuse regarding the use of echocardiograms after valve replacement reported in our study (59.6%) was significantly greater than that reported in a Canadian study conducted from 2014 to 2016 (20.0%) [29], although the latter included a broader range of indications that were not exclusively surgical. While overuse usually has a multicausal origin [30], the fear of potential complications associated with the valve replacement surgery (such as acute myocardial infarction, haemorrhage or cardiac conduction abnormalities [31]) may have contributed to the over-indication of echocardiograms. In this context, the accessibility, low cost, and non-invasive nature of this diagnostic technique could be related to a lower perceived risk of overuse [3].

Nonetheless, the low degree of overuse identified for cholecystectomies (2.0%) was very similar to that reported in Spain in 2000 (0.7%) [32], confirming a downward trend compared to other studies performed between 1986 and 1991 [33, 34], which applied criteria now considered obsolete [35]. A similar situation was observed for Mohs surgery, where the absence of overuse aligns with the 2.0% reported by the University of Utah (United States), which even suggests a potential underuse of this technique [36]. This study is the first to address the overuse of hallux valgus/ hammer toe interventions and the perioperative use of opioids in pediatric patients, limiting the ability to compare this data. However, previous studies have documented an opioid overuse rate of 26.0% in the general population [37].

For these three previous procedures (cholecystectomies, perioperative use of opioids in pediatric patients and podiatric surgery), the low degree of overuse reported in this and other studies could call to question the validity of their respective recommendations. In this regard, these findings are contrary to the notion that the “Choosing Wisely” recommendations should be directed to unnecessary practices that are frequently performed in our environment [38]. Therefore, although these recommendations may have fulfilled this premise at the time of their development, the ongoing absence of overuse observed in these scenarios (and consequently, the inability to achieve further reduction of inappropriateness) would limit the potential benefits and impact on current clinical practice.

On the other hand, the higher risk of overuse reported in elderly patients, in those with ≥ 1 IRF, with ASA grade III-IV or an intermediate Charlson index is also consistent with the scientific literature [32, 39, 40]. In this context, various studies have linked the presence of comorbidities to a higher risk of complications resulting from surgical procedures [32, 41], as well as to a lower level of consensus regarding the appropriateness of the procedure indication [3]. Moreover, patients who were referred to specialized care upon discharge showed a higher frequency of overuse compared to those who continued their follow-up in primary care. This persistence of care complexity may be associated with defensive medicine practices, where healthcare professionals aim to maintain a greater control and security in managing the clinical case. Defensive medicine is, in fact, considered one of the main causes of overall overuse, encompassing surgical practices [42]. In addition, evidence suggests that cognitive biases, established clinical culture, patient expectations, and healthcare system incentives play an important role in perpetuating inappropriate use of medical and surgical interventions [43, 44].

The cost attributed to overuse was estimated at €86,858.05. Most of these expenses originated from the overuse of surgical interventions due to the methodology used and the foreseeable higher unit cost of these procedures. If these costs were extrapolated to the 70 hospitals with more than 500 beds existing in Spain, the total would amount to 6,080,063.5 euros [45]. However, comparing these results with data from other studies is difficult, as the prices of the procedures of interest may differ among healthcare systems and countries, and change over time due to inflation or the implementation of new technologies [46].

A number of limitations of this study should be noted. (1) The opportunistic selection of the evaluated recommendations does not allow generalization of the results to the entire surgical practice. However, we attempted to minimize this risk by selecting several recommendations representative of different surgical settings. (2) There is a possibility that some procedures of interest were not identified due to the incorrect assignment of a DRG code. However, the use of systematic sampling based on the hospital’s coding systems allowed for a representative identification of clinical episodes in which these interventions were performed. Nevertheless, this would result in an underestimation of the absolute overuse values and associated costs without affecting the obtained degree of overuse; (3) Of the selected recommendations, only the one related to cholecystectomies in patients with asymptomatic cholelithiasis is equivalent to the indication established by the Spanish Association of Surgeons [47]. However, the other “Choosing Wisely” recommendations selected are still aligned or compatible with the indications issued by various Spanish and European scientific societies [48–51]; (4) The data collection period was 2018–2019; therefore, some variation in the results compared with the current situation cannot be excluded.

The strengths of this study include the following. First, this is the first study to estimate the degree of overuse regarding the perioperative use of opioids in pediatric patients and surgery for hallux valgus/hammer toe. Second, the study employs an integrated epidemiological approach combining patient IRF, intervention-related characteristics, and cost analysis to evaluate overuse in surgery and to quantify its direct economic burden.

Conclusions

Applying five “Choosing Wisely” recommendations, overuse was identified in 9.7% of cases and was mainly concentrated in echocardiograms performed after valve replacement, podiatric surgery, and, to a lesser extent, cholecystectomies. No evidence of overuse was found for Mohs surgery or for the perioperative use of opioids in pediatric patients. Overuse was more frequent among older patients, those with one or more IRF, and interventions performed in hospitalized patients or within scheduled surgical care. The direct economic impact associated with overuse exceeded €86,000. These findings highlight specific procedures and care settings where the systematic application of “Choosing Wisely” recommendations could support the reduction of low-value care while improving patient safety and resource use.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ABIM

American Board of Internal Medicine

ASA

American Society of Anesthesiologists Physical Status scale

DRG

Diagnosis-Related Group

IFR

Intrinsic risk factors

IQR

interquartile range

95% CI

95% confidence intervals

Author contributions

JMAA, JVG, and AGSV contributed to the conceptualization and methodology of the study. AGSV, JVG, and DSJS performed the formal analysis and investigation. AGSV and JVG wrote the original draft of the manuscript. DSJS, PS, CDAP, and JMAA contributed to the review and editing of the manuscript. All authors read and approved the final manuscript.

Funding

The study received funding for the electronic medical records review from project FIS PI 16/00971 and its translation and dissemination through the competitive call of the International University of La Rioja: ‘III CONVOCATORIA DE PROYECTOS DE INVESTIGACIÓN - FACULTAD DE CIENCIAS DE LA SALUD’.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This project was approved by the Research Ethics Committee of the Ramón y Cajal University Hospital (reference 168/17, dated 07/31/2017) and by the International University of La Rioja (reference PI015/2023, dated 03/07/2023). The research was conducted in accordance with the ethical principles set forth in the Declaration of Helsinki.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.