Sonography After Thoracic Surgery (SATS)

Following thoracic surgery (both open and thoracoscopic) patients have a surgically induced pneumothorax and hence tube thoracostomy is needed to drain the air and effusion. Due to this, patients must undergo post operative chest x-ray (CXR) evaluations in order to evaluate the chest and make decisions regarding removal of chest tubes as well as for decisions regarding patient discharge. Following chest tube removal, most surgeons advocate further CXR evaluation to assure that no pneumothorax was introduced during removal of the tube.

Repetitive CXR evaluations are costly, inconvenient, expose the patient and healthcare worker to repetitive radiation, and require time. Surgeon-performed US could be performed as an adjunct to physical examination. This would allow surgeons to make decisions regarding chest tube removal during rounds, alleviate the need for most CXRs as well as decrease patient length of stay. The goals of this project is to: evaluate the reliability and accuracy of surgeon-performed US following thoracic surgery, and evaluate the potential improved convenience for patients and the healthcare system using surgeon-performed US.

US technology has traditionally been the domain of the radiologist; however, recently US has evolved into an important tool for the clinician, similar to the stethoscope of the past, which can be utilized in an organ and specialty specific manner. Users of US in clinical care on a daily basis for specialty specific indications are: cardiologists, gastroenterologists, anesthesiologists, OBGYN, trauma, breast, ENT, hepatobiliary, and colorectal surgeons. US technology has yet to find its place in the domain of the thoracic surgeon and in the diagnosis and treatment of pulmonary malignancy, although there are many potential important applications for its use. We believe that portable hand-held ultrasound may be the thoracic surgeon's stethoscope of the future.

Thoracic ultrasound has been shown to be accurate at diagnosing pneumothorax (sensitivity, specificity = 90%) and has been well studied in the trauma population (sensitivity 95%) , however, few studies have been done using thoracic ultrasound in the non-injured population . To our knowledge, there are currently no centers using thoracic ultrasound routinely in the post-operative setting following general thoracic surgery.

Buzzas et al. showed that surgeons and surgical residents performing focused abdominal sonography for the initial evaluation of trauma patients were able to correctly diagnose injuries as safely and accurately as radiologists . Sensitivity, specificity, accuracy and negative predictive values were not significantly different between the surgeon and radiologist groups.

This study will consist of a prospective evaluation of surgeon-performed US following thoracic surgery. CXR diagnosis of pneumothorax/pleural effusion will be taken to be the "gold standard" and US detection of pneumothorax/pleural effusion will be measured against CXR. Standard post-operative evaluation and management of these patients will be applied. Thoracic US evaluation of the chest will occur in the same time sequence as CXR for all patients enrolled in the study. The results of the US will be compared with the CXR findings. Proposed management decisions based on US will be compared with the actual management decision made by the surgeon based on the CXR finding.

The proposed project is a clinical feasibility and safety study concerning a new indication for surgeon performed ultrasound in general thoracic surgery. Surgeon performed ultrasound has recently become popular in the initial evaluation of the injured patient as well as for office-based visualization of cysts and masses. It is also being used to aid in biopsy of masses and attaining central venous access in the critical care setting.

We plan to evaluate surgeon performed ultrasound in the evaluation of post-operative pneumothorax and pleural effusion following thoracic surgery. Surgeon performed ultrasound will be compared to plain film CXR, which is considered the "gold standard". If the results of this study show that surgeon performed ultrasound is accurate and sensitive in the evaluation of post-operative thoracic surgery patients, ultrasound could replace CXR in the post-operative care of these patients.

Ultrasound evaluation of the pleural space by surgeons post-operatively would lead to decreased costs, length of stay and patient inconvenience. Surgeon-performed ultrasound could be used as an adjunct to physical examination during ward rounds. This would not only make the decisions regarding chest tube management possible during morning rounds, it would eliminate the cost of repeated CXRs. Following the initial cost of purchasing the ultrasound machine, performing the test on a daily basis would not cost anything. Furthermore, patients (with chest tubes) would no longer need to be transported to the radiology department in order to undergo proper evaluation of the pleural space.

There is a tremendous advantage to minimally invasive diagnostic and treatment strategies such as the ones described in this proposal to healthcare systems. By enabling early, rapid, and safe diagnostic and treatment strategies for patients, healthcare systems benefit from shorter hospital duration of stays, waiting times for procedures and costs. This is extremely attractive and important in a Canadian environment. This project will lead to enhanced process of care for lung cancer patients.

Hypotheses:

Bedside transthoracic ultrasonography is a simple, quick, reproducible, cost-effective and novel method of accurately evaluating post-operative thoracic surgery patients.

Objectives:

• Describe the surgeon and resident ability to learn and apply thoracic ultrasound to an inpatient population.
• Evaluate the reliability of surgeon-performed ultrasound following general thoracic surgery.
• Compare the criterion-related validity, and specifically the sensitivity, specificity, positive and negative predictive value of surgeon-performed ultrasound at diagnosing post-operative pneumothorax and effusion using chest x-ray as the criterion or "gold standard".
• Evaluate the marginal cost minimization associated with using surgeon performed ultrasound post thoracic surgery compared to conventional chest x-ray.

Study Design:

Prospective, cohort study design. Clinical Trial: Intervention - pleural US. All patients will receive post-operative pleural-based ultrasound during their follow-up care on the thoracic surgery ward.

Methods:

This study will consist of a prospective non-randomized evaluation of surgeon-performed ultrasound following thoracic surgery. All patients undergoing elective, open and thoracoscopic thoracic surgery for any indication at the CHUM - Hopital Notre Dame, Montreal, Quebec, Canada will be approached in an attempt to enrol them into the study. Patients will be enrolled and consented for the study by the treating physician during the pre-operative evaluation in the pre-operative clinic.

There are approximately 12 potential patients per week. Chest x-ray (CXR) diagnosis of pneumothorax will be taken to be the "gold standard" and ultrasound detection of pneumothorax will be measured against CXR. The sensitivity of ultrasound in diagnosing pneumothorax has been shown to be 95% .

A clinical nurse specialist will acquire all required data from interviews with each patient as well as from each patient's chart. He or she will enter all data into a prospective database specifically designed for the study. The nurse will also be responsible for assuring that the surgeons complete study case report forms correctly after each ultrasound evaluation. He or she will also record the results of all CXRs into the database.

Standard post-operative evaluation and management of these patients will be applied during the entire study. Chest x-rays will be ordered as usual and decisions regarding chest tube removal and management will be made based solely on CXR findings. Thoracic ultrasound evaluation of the chest will occur in the same time sequence as CXR for all patients enrolled in the study. A thoracic surgeon or senior surgical resident trained in thoracic ultrasound will perform all ultrasounds. The surgeon will record the outcome of his/her evaluation of the chest as a dichotomous variable (pneumothorax - YES/NO, pleural effusion - YES/NO) on a specialized study case report form. The surgeon will also record a management decision on the study report form (the management that he/she would have undertaken if he/she was acting on the ultrasound finding alone). The results of the ultrasound will be compared with the CXR findings (as reported by staff radiologists). Proposed management decisions based on ultrasound will be compared with the actual management decision made by the surgeon based on the CXR finding. The radiologist will be blind with respect to the surgeon's ultrasound assessment and vice-versa.

Learning Phase During the first two weeks of the study, surgeons and residents will practice the ultrasound techniques on 24 patients. The results of their findings will be compared to CXR and will be evaluated in order to assess the learning curve for this technique. During this phase of the study, any discrepancy between the surgical staff's findings and the CXR will be reported to the staff during the same day to provide feedback that will enhance the learning process. After the surgical staff reviews the CXR, he or she will then re-evaluate the patient using the ultrasound machine in order to learn from his or her mistake. Management decisions regarding patient care based on the ultrasound will not be recorded during the "learning phase" of the study.

Assessment Phase Following the learning phase, the assessment study will commence. During this phase, all patients enrolled in the study will undergo ultrasound evaluation of the pleural space every time a CXR is ordered in the post-operative period. The ultrasound for each patient will be conduced by two different surgeons that will be selected randomly or pseudo-randomly based on availability.

This phase of the study will be conducted in order to define the reliability of ultrasound in diagnosing pneumothorax/pleural effusion, as well as the ability of ultrasound to be used as the sole decision-making diagnostic test following general thoracic surgery.

Reliability and specifically inter-rater reliability will be assessed by the degree of agreement of the ultra-sound results recorded by the two surgeons. The kappa statistic will be used to assess the degree of agreement with a minimal kappa level of 70%.

Sensitivity, specificity, positive predictive value, negative predictive value, overall accuracy and likelihood ratios will be computed for the surgeon-ultrasound against the CXR. Each surgeon's observation will be considered as an independent record, thus removing the complication of disagreement between surgeons and possible confounding. The psychometric properties listed above will be estimated as crude values. However, subgroup analyses and logistic regression will evaluate the impact of surgeon/assessor, learning curve slope of the surgeon, difficulty of case as assessed by the radiologist on the accuracy and the likelihood of false negative and false positive results.

Marginal costs differences between the surgeon performed ultrasound and CXR will be based on an overall risk-benefit assessment that will take into account training requirements, accuracy and costs of facilities as well as physician services. Cost assessment will take into account the management of potential false negative and false positive results.