Elimination of routine chest X-rays after lung resection: a quality improvement initiative

Introduction

Chest X-rays (CXRs) are routinely and widespread performed after lung surgery, with practices varying widely between centers. In the absence of clear clinical indications, routine CXRs are often performed immediately after surgery, at regular intervals, after drain removal, or even preoperatively (1). Several reasons may explain this practice, such as monitoring for complications or documenting good postoperative progress. However, it is likely more a tradition that has been passed down over time.

Although CXRs are low-radiation procedures, their routine use postoperatively remains a topic of debate. Beyond the financial costs, unnecessary CXRs add logistical and cognitive burdens on healthcare providers and identifying incidental findings with no clinical significance can lead to overtreatment and associated complications. Several studies have questioned the clinical utility of routine CXRs after lung surgery. For instance, Porter et al. found that while 48% of patients had abnormal findings in postoperative CXRs, only one case led to a minor change in care, and none required intervention (2). This study questioned the utility of routine postoperative CXRs but did not analyze the effect of eliminating them on patient outcomes. Similarly, Bjerregaard et al. demonstrated that only 4% of CXRs after thoracoscopic surgery revealed abnormalities, with less than 1% leading to clinical interventions (3). Alcasid et al. demonstrated that an abnormal CRX at day 0 was not predictive of procedural requirement (4). A systematic review also showed that routine CXRs after surgery led to changes in clinical management in only 2.8% of immediate postoperative CXRs and 7.3% of post chest tube removal CXRs, most often resulting in repeat imaging rather than significant treatment changes (1).

Despite this evidence, only one study has investigated the impact of discontinuing routine CXRs on clinical outcomes, and none has conducted a before-after analysis of this intervention (5). This initiative aims to address this gap by evaluating the safety and efficacy of discontinuing routine CXRs after lung resection, with a focus on readmission rates, cardiopulmonary complications, inpatient mortality, and hospital length of stay. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-911/rc).

Methods

Patient

This study is a single-center, retrospective, controlled before-after study. It included all patients who underwent elective pulmonary lobectomy, segmentectomy and wedge resection in the Division of Thoracic and Endocrine Surgery, Geneva University Hospitals and Faculty of Medicine, Geneva, Switzerland, between January 2022 and May 2023. Patients treated for pneumothorax, pneumonectomy, or lung volume reduction surgery were excluded. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Cantonal Commission for Ethics and Research (CCER) of Geneva (referral number: 2024–00628) and the need for informed patient consent was waived.

Initiative

The initiative aimed to discontinue routine CXR, which were previously performed postoperatively upon leaving the post-anaesthetic care unit (day 0), after chest drain removal, and during postoperative follow-up between 10 and 14 days. CXRs were only prescribed when clinically indicated, such as in cases of unexpected desaturation, new onset dyspnoea, prolonged (>7 days) or significant air leaks, or clinical signs of pneumonia. The initiative started on December 1, 2022, and data were collected for six months until May 31, 2023. Comparisons were made with a non-matched control group of patients consecutively selected between January and November 2022 (Figure S1). Data for the control group were collected retrospectively. Additional analyses were conducted 10 months post-initiative to monitor practice evolution.

Clinical outcomes and definitions

Medical records were extracted from the hospital’s data management system, analysing patient demographics, type of lung resection, surgical approach, and indications for surgery and CXRs. Diagnosis of chronic obstructive pulmonary disease was documented without stratifying phenotypes. Non-active smoker was defined as no active smoking in the 4 weeks prior to surgery. All radiological data were obtained from the official radiology reports documented in the hospital’s electronic medical record system. The interpreting radiologists were not involved in the study and were unaware of the study hypothesis or group allocation at the time of reporting. Abnormal CXR findings from radiological reports were classified as pleural effusion/hemothorax, pneumothorax, mediastinal emphysema, elevated hemidiaphragm, pulmonary consolidation, atelectasis, and extra pleural positioning of the chest tube. Any change in patient care due to CXR findings, including thoracic drainage, reoperation, or additional radiological checks, was noted.

The primary objective was to evaluate the impact of discontinuing routine CXRs on the 30-day readmission rate to the same center. The rationale is that missed complications undetected on routine CXRs in asymptomatic patients may result in rehospitalization if further intervention is required, including during the outpatient consultation. Therefore, patient safety in the early postoperative period in asymptomatic patient was expected to remain unchanged with the new approach.

Secondary outcomes included chest drainage duration in days, inpatient mortality in %, cardiopulmonary complication rate in %, reoperation rate, length of stay in days, and 30- to 90-day readmission rates to the same center. Cardiopulmonary complications included atrial fibrillation, myocardial infarction, pneumonia, pulmonary embolism, pneumothorax, hemothorax, prolonged air leak (PAL), and acute respiratory distress syndrome (ARDS). Pneumonia was defined as the need for antibiotics based on radiological lung changes, fever, or an elevated white blood cell count above 12,000/mL. PAL was defined as an air leak lasting beyond postoperative day 7, while ARDS was diagnosed according to the Berlin definition based on Ferguson et al. (6).

Follow-up analysis from August to September 2023 to assess the evolution of practice focused on the 90-day readmission rate and the frequency of postoperative CXRs.

Statistical analysis

Nonparametric continuous variables were presented as median with interquartile range (IQR) and normally distributed continuous variables as mean with standard deviation. Categorical and ordinal variables were presented as numbers with percentages. A Chi-squared or Fisher’s exact (if expected cell counts ≤5) test was used to analyse categorical variables. A t-test or Mann-Whitney U test was used to compare continuous variables, for parametric and non-parametric variables respectively. All analyses were performed using STATA software, version 14 (StataCorp LLC, Texas, USA).

Results

A total of 128 patients were included during the study period, with 64 patients in each group. Both groups were comparable in terms of age, gender distribution, comorbidities, surgical indication, type of resection performed, proportion of video-assisted thoracoscopic surgery (VATS) approach, and the percentage of patients leaving the operating room with a chest drain (Table 1). Nine out of the eleven patients who did not have a chest drain postoperatively underwent a wedge resection.

CXR

Overall, the number of CXR decreased from 247 before the initiative to 109 CXRs after the initiative. The median number of CXRs per patient decreased from 3 (IQR, 3–5) to 1 (IQR, 0–3) (Table 2, P<0.001). As expected, there was a significant reduction in the number of prescribed CXRs on day 0 (P<0.001), after chest tube removal (P<0.001), and during outpatient follow-up (P<0.001) following the initiative. The number of CXR performed during these three periods decreased from 186 to 54. Among these 240 CXR in total, 126 were reported as abnormal by radiologists, leading to an additional invasive procedure in only 3 patients (2.3%).

Only 28.1% of patients had no CXR after the initiative. However, in this group, 73.9% of CXRs on day 0, 30.7% after chest drain removal, and 5.6% during outpatient follow-up were ordered by mistake (Table 3).

The rate of abnormal CXR and the need of supplementary invasive procedure were similar between groups. Overall, the CXR performed on day 0 immediately after the surgery is abnormal in 37.9% of cases and no patient led to an invasive procedure (Table 2). After chest tube removal, CXRs were abnormal in 74.3% of patients and only 2 patients required a new chest drain. During the outpatient follow-up, 61.7% of CXR were abnormal and only 1 patient required a hospitalisation.

The rate of patients who received a CXR during hospitalisation outside routine indications was comparable between groups (Table 2). These CXRs have the highest rate of abnormal findings (87.8%), the highest rate of change to clinical management (65.8%) and the higher rate of patients requiring a new invasive procedure (7.3%), compared to routine CXRs or CXRs performed at routine times after the start of the initiative.

Clinical outcomes

The 30-day readmission rates in the group after and before the start of the initiative were 4.8% (n=3) and 14.1% (n=9) respectively (P=0.07) (Table 4), the complication rates 23.4% (n=15) and 30.1% (n=19) (P=0.39), the reoperation rates 3.13% (n=2) and 0% (n=0) (P=0.50), the mortality rates 1.6% (n=1) and 0% (n=0) (P>0.99), and the 30-to-90-day readmission rates 1.6% (n=1) and 3.1% (n=2) (P>0.99). The indication for the two readmissions to theatre was post-operative bleeding in both cases and the cause of death ARDS following pneumonia, thus not linked to the discontinuation of CXR and clinically identifiable without routine CXR. The median duration of drainage was 1 day (IQR, 1–3 days) in the after group and 1 day (IQR, 1–2 days) in the before group and the median length of stay was 3 days (IQR, 2–5 days) in both groups (Table 4).

Follow-up analysis at 10 months

The results of the follow up analysis at 10 months of the implementation of the initiative are shown in Figure 1. Twenty-two additional patients were included over a period of 2 months (Table S1). The percentage of patients without any CXR rose from 28.1% to 77.2% (n=17) and the readmission rate fell from 4.8% to 0%.

Figure 1 Changes in the percentage of patients without postoperative CXRs and the readmission rate across the three analysed periods: before the initiative, immediately after its implementation, and 10 months post-implementation. CXR, chest X-ray.

Discussion

We conducted an exploratory, before/after study evaluating the impact of discontinuing routine CXRs after lung surgery on 30-day readmission rate and other complications during the inpatient postoperative phase. To the best of our knowledge, this is the first study to report a prospective evaluation of patient outcomes after discontinuing routine postoperative CXRs and compare it to historical controls within the same institution. Our results suggest that discontinuing routine CXRs after lung resection surgery (excluding patients treated for pneumothorax, pneumonectomy, or lung volume reduction surgery) is safe and does not lead to increase re-admission or complication rates.

The analysis of CXR results revealed that a significant proportion of the routine CXRs performed postoperatively showed abnormalities; however, their influence on patient care was limited which is consistent with existing literature (1,7,8). Among the CXRs conducted immediately post-surgery, after chest drain removal, and during follow-up consultations, only a small percentage led to actionable clinical decisions. For instance, abnormal imaging findings were often minor or incidental, such as small effusions or atelectasis, which did not necessitate further intervention. In cases where clinical action was required, the interventions were typically minimal, such as ordering repeat imaging or extending observation time rather than more invasive procedures like chest drainage or reoperation.

CXRs ordered on a non-routine basis during the hospital stay showed the highest rate of abnormal findings and subsequent changes in clinical management. This outcome is expected, as these CXRs were typically requested in response to clinical presentations that deviated from the anticipated postoperative course. As already discussed by other authors (9-11), the decision to change clinical management is indeed rarely based on radiological findings alone and include additional clinical information. Therefore, the decision to perform a CXR should be primarily driven by a clinical presentation, not routine. In our study, the three major complications observed, two re-operations and one death, had clinical presentations identifiable without routine CXRs.

A previous systematic review showed that chest ultrasound could significantly reduce the use of CXR in postoperative care after surgery (12). However, we did not include or assess chest ultrasound in our study. While ultrasound shows promise, there is a potential risk of overtreatment if it becomes a routine practice. It can also be time-consuming, and interpreting results may be challenging without a standardised methodology. Its use for specific clinical indications requires further evaluation and comparison to ensure its appropriate application. Beside clinical evaluation, the current postoperative management of patients who underwent lung resection also increasingly includes digital drainage systems (13), whose outputs give important information about the postoperative trajectory and offer an alternative to CXRs in the detection of deviation from the norm.

The absence of routine CXRs did not significantly impact patient outcomes, with no increase in the rate of serious complications or missed diagnoses. Additionally, the slight reduction in the number of complications observed may be explained by a lower detection rate. Similarly, the minor reduction in the 30-day readmission rate could potentially be due, in part, to less overtreatment of isolated radiological findings during follow-up consultations. This suggests that in most cases, routine CXRs provided little additional value in guiding postoperative care, reinforcing the idea that CXRs should be reserved for cases with clear clinical indications. The stable readmission rates between 30 and 90 days in both groups suggest that a lower 30-day readmission rate does not negatively affect long-term outcomes. Although our study did not observe an impact on the length of stay, Haddad et al. previously reported a reduction in length of stay for patients who did not undergo routine CXRs, which may be attributed to similar mechanisms (5).

An important observation from our study was the persistence of erroneous CXR prescriptions during the early phase of implementation. These “mistakes” were primarily attributable to ingrained habits among healthcare staff, particularly in the post-anesthesia care unit, where automatic prescriptions continued out of routine. This highlights a common and well-documented challenge in quality improvement initiatives: behavioral change often lags behind protocol updates, especially when longstanding practices are being reconsidered. However, the significant reduction in unnecessary CXRs observed in the 10-month follow-up period demonstrates that adherence improved over time. This shift in clinical practices, showing that the surgeon’s growing confidence, along with the increased familiarity of the entire care team with the new approach. This emphasizes the importance of repeated communication, education, and reinforcement when implementing change (14). The continued decrease in both the complication and readmission rates over time suggests that overtreatment is being avoided, likely due to fewer unnecessary interventions and more selective use of diagnostic imaging. This evolution indicates that the practice changes not only maintained patient safety but also promoted more efficient and targeted care.

In addition to the benefits for patients, such as reduced exposure to ionising radiation, and for healthcare professionals, by easing logistical and cognitive workloads, discontinuing routine CXRs also offers financial advantages for the healthcare system. Our preliminary cost analysis revealed a potential savings of 14,000 CHF (14,910 EUR as of 01.01.2025) over five months at our center. If unnecessary CXRs, performed by mistake or for routine follow-up, were eliminated, with a 69% reduction in CXRs achieved, the savings could rise to 17,100 CHF (18,212 EUR) over five months, equating to 41,040 CHF (43,708 EUR) annually. Furthermore, this estimate does not yet include the potential cost savings from avoiding re-admissions, which could be realised if the trend observed in our study is validated by future research.

The results of this study should be interpreted in the context of several limitations. First, the data analysed focus on the early stage of implementation, were several CXRs were still prescribed by mistake. Part of the results, and the absence of difference in the primary and secondary outcomes between the two groups, could therefore be explained by the fact that over a third of the patients in the after group received a similar treatment that the patient in the before group. However, a statistically significant difference in the number of CXRs per patient was observed and just under a third of the participants in the after group had no CXRs at all. Moreover, the follow up analysis showed similar trends in term of 30-day readmission rates in a group with a much higher percentage of patient not receiving any CXRs at all. Second, no correction for multiple comparisons was implemented. However, the aim of this study was not to prove the existence of significant differences between the two groups, but on the contrary to show that discontinuing routine CXRs did not increase readmission or complication rates. Third, several care decisions impacting the study implementation or outcomes were based on clinical judgement, which is not entirely reproducible. Although such decisions, like performing a non-routine CXR or readmit a patient, were based on commonly accepted clinical cues, they were not strictly protocoled.

All but two of our patients were operated by VATS, which limits the generalization of our results to patients undergoing open lung resections. Fourth, as this was an exploratory quality improvement initiative rather than a hypothesis-testing study, no formal sample size calculation was performed beforehand. However, our primary objective was to assess whether discontinuing routine CXRs would compromise patient safety. A lower readmission rate in the post-intervention group supports the hypothesis. Finally, this study was not hypothesis testing and its results should therefore not be interpreted as formal evidence of non-inferiority. Likewise, the trends observed in some of the outcome, such as the decreased readmission and complication rates, should be interpreted as hypothesis generating for further studies rather than proves of differences.

To minimize selection bias, we included all consecutive patients meeting inclusion criteria over the study period. The observer bias was minimized by relying on objective endpoints (readmission, reoperation, mortality) and by extracting radiology results from standardized radiology reports.

Conclusions

In summary, while routine CXRs occasionally detected abnormalities, their overall impact on patient care was minimal, including during the outpatient postoperative follow-up consults. The shift to clinically indicated CXRs demonstrated that eliminating routine imaging did not compromise patient safety and reduced unnecessary interventions. Beyond the influence on direct patient outcomes, discontinuing unnecessary tests could also be beneficiary to the healthcare systems more broadly, reducing costs and workload on already strained resources, although this would need formal validation by future studies.

Acknowledgments

The abstract of this article was presented in the 32nd Meeting of the European Society of Thoracic Surgeons (2024).

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-911/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-911/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-911/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-911/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Cantonal Commission for Ethics and Research (CCER) of Geneva (referral number: 2024–00628) and the need for informed patient consent was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Galata C, Cascant Ortolano L, Shafiei S, et al. Are Routine Chest X-rays Necessary following Thoracic Surgery? A Systematic Literature Review and Meta-Analysis. Cancers (Basel) 2022;14:4361. [Crossref] [PubMed]
2. Porter ED, Fay KA, Hasson RM, et al. Routine Chest X-Rays After Thoracic Surgery Are Unnecessary. J Surg Res 2020;250:188-92. [Crossref] [PubMed]
3. Bjerregaard LS, Jensen K, Petersen RH, et al. Routinely obtained chest X-rays after elective video-assisted thoracoscopic surgery can be omitted in most patients; a retrospective, observational study. Gen Thorac Cardiovasc Surg 2015;63:465-71. [Crossref] [PubMed]
4. Alcasid NJ, Banks KC, Jiang SF, et al. Are routine, daily chest radiographs (CXR) necessary following (VATS and RATS) lobectomies? Surg Open Sci 2024;20:20-6. [Crossref] [PubMed]
5. Haddad L, Bubenheim M, Bernard A, et al. The Interest of Performing "On-Demand Chest X-rays" after Lung Resection by Minimally Invasive Surgery. Thorac Cardiovasc Surg 2017;65:572-80. [Crossref] [PubMed]
6. Ferguson ND, Fan E, Camporota L, et al. The Berlin definition of ARDS: an expanded rationale, justification, and supplementary material. Intensive Care Med 2012;38:1573-82. [Crossref] [PubMed]
7. Thet MS, Han KPP, Hlwar KE, et al. Efficacy of chest X-rays after drain removal in adult and pediatric patients undergoing cardiac and thoracic surgery: A systematic review. J Card Surg 2022;37:5320-5. [Crossref] [PubMed]
8. Dezube AR, Deeb A, De Leon LE, et al. Routine Chest Roentgenogram After Chest Tube Removal Is Not Indicated for Minimally Invasive Lung Resection. Ann Thorac Surg 2022;114:2108-14. [Crossref] [PubMed]
9. Whitehouse MR, Patel A, Morgan JA. The necessity of routine post-thoracostomy tube chest radiographs in post-operative thoracic surgery patients. Surgeon 2009;7:79-81. [Crossref] [PubMed]
10. Zukowski M, Haas A, Schaefer EW, et al. Are Routine Chest Radiographs After Chest Tube Removal in Thoracic Surgery Patients Necessary? J Surg Res 2022;276:160-7. [Crossref] [PubMed]
11. Cerfolio RJ, Bryant AS. Daily chest roentgenograms are unnecessary in nonhypoxic patients who have undergone pulmonary resection by thoracotomy. Ann Thorac Surg 2011;92:440-3. [Crossref] [PubMed]
12. Grapatsas K, Leivaditis V, Ehle B, et al. Can Chest Ultrasound Replace Chest X-ray in Thoracic Surgery? Tomography 2022;8:2083-92. [Crossref] [PubMed]
13. Chang PC, Chen KH, Jhou HJ, et al. Promising Effects of Digital Chest Tube Drainage System for Pulmonary Resection: A Systematic Review and Network Meta-Analysis. J Pers Med 2022;12:512. [Crossref] [PubMed]
14. Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients' care. Lancet 2003;362:1225-30. [Crossref] [PubMed]