2.1. Study Design and Patients

2.2. Inclusion and Exclusion Criteria

Intra- and postoperative exclusion criteria were as following: intraoperative conversion to open surgery due to adherences or major complications (bleeding > 2000 mL, anesthesiologic complications requiring reintubation after surgery or surveillance in the intensive care unit), the occurrence of early postoperative complications needing hospital monitoring or redo-operation, difficult pain management (Numerical Rating Scale > 7), the occurrence of a postoperative complication preventing a safe discharge (i.e., persistent air leaks with incomplete lung re-expansion), a peripheral oxygen saturation <92% requiring O2 supplementation, altered vital or laboratory parameters (i.e., a systolic blood pressure <95 or >160 mmHg, body temperature >37 °C, heart rate >100 bpm).

2.3. Protocol

The telehealth home monitoring protocol and the function of the device were explained during surgical and anesthesiologic outpatient visits systematically performed before hospital admission, thus offering pre-operative counselling to reduce fear and anxiety regarding their role in the study. All enrolled patients started respiratory physiotherapy at least one week before surgery thanks to multimedia information containing explanations for procedures downloaded by a QR code. All surgical procedures were carried out under general anesthesia with a Da Vinci Xi system (Intuitive Surgical, Sunnyvale, CA, USA). A lung-protective ventilation technique was adopted (tidal volume 4–6 mL/kg of predicted body weight, positive end-expiratory pressure–PEEP between 5 and 8 cm H₂O, fraction of inspired oxygen between 0·5 and 0·8). An anesthesiologic multimodal protocol was adopted for all patients. This consisted in the execution of a preoperative erector spinae plane (ESP) block (levobuvicaine 0·25% 20 mL), administration of dexamethasone 4–8 mg, sulphate magnesium 1 gr before surgical incision and ketorolac 30 mg and paracetamol 1 g 30 min before awakening. In the postoperative period, pain was controlled by continuing non-steroidal anti-inflammatory drugs (ibuprofen 600 mg twice daily) and paracetamol (1000 mg three times daily). At the end of surgery, pleural drainage was ensured by placing a single 28 Fr chest tube. A Heimlich valve was systematically placed on the afternoon of POD 1. The chest tube was removed before discharge when the daily pleural effusion was less than 200 cc and no air leakage was observed.

In case of air leakage, the chest tube was left in place and the discharge was only allowed in case of adequate lung re-expansion (if the lung reaches the thoracic wall) on chest X-ray.

2.4. Telehealth Home Monitoring

Before inclusion in the study, all patients and their caregivers were informed on the need for the use of the device during the study period (from discharge to removal of chest drain). The device was delivered to patients the day before surgery and each patient was trained in the presence of their caregiver during hospitalization to avoid emotional stress affecting their understanding of how to correctly use the device. All clinical data recorded during the preoperative training period and the first POD was excluded from the formal analysis of this study.

Teleconsultations were scheduled as follows: first, one in the afternoon of POD2 once the patient reached his home after discharge, twice on POD3 (one in the morning and one in the afternoon), and then once a day until chest tube removal in an outpatient visit. The criterion for chest tube removal was the presence of a daily output lower than 200 cc in absence of air leakage.

After discharge, patients self-recorded their vital signs (blood pressure, body temperature, heart rate, and peripheral oxygen saturation) at least four times a day using the device. These readings were instantly accessible for review by two surgeons (research physicians) through a dedicated smartphone application. The patient was also allowed to visualize the reordered parameters.

In the event of any anomalies in the recorded vital signs (threshold violations), an alert was sent directly to the research physicians via mobile phone, including both text messages and emails.

Threshold violations were classified as moderate (yellow) and critical (red). Yellow threshold violations included the following: systolic blood pressure > 140 or <100 mmHg, diastolic blood pressure > 95 or <60 mmHg; oxygen saturation < 94%; heart rate > 100 bpm; and temperature > 37.5 °C. Red threshold violations were defined as follows: systolic blood pressure > 160 or <80 mmHg, diastolic blood pressure > 110 or <40 mmHg; oxygen saturation < 89%; heart rate > 140 bpm; and temperature > 38 °C.

In the event of red threshold violations, research physicians were alerted, prompting immediate contact with patients by telephone to obtain further information about parameter deviations and their clinical status. This procedure was also implemented in cases of missing data to provide technical assistance to the patients whenever needed.

Additionally, a direct telephone line was available to patients 24 h a day, along with a dedicated re-hospitalization pathway in the event of sudden changes in vital signs or the occurrence of adverse events. Lastly, all patients were monitored according to the standard of care following the removal of the chest drain. For all enrolled patients, a 30-day follow up was performed.

2.5. Outcome Measures

Primary outcomes were: (a) feasibility (rate of successful enrolment and completion of the protocol), and (b) safety (assessed through the occurrence of post-discharge complications and the number of hospital readmissions).

Secondary outcomes were: (a) comparison of results with matched controls in which the standard of care was maintained, and (b) economic evaluation (in terms of hospitalization days and costs saved).

2.6. Statistical Analysis

All data were collected and stored in a Microsoft Excel^® spreadsheet. Descriptive statistics were employed to present baseline and surgical characteristics of patients. Continuous variables are reported as mean and standard deviation, while discrete variables are reported as counts and percentages. A matched control group for safety evaluation of this protocol was obtained from a clinical cohort of patients who underwent robotic lobectomy for stage I–II NSCLC in our Department between 2019 and 2023. Each case was matched with a minimum of 1 to a maximum of 3 controls (35 cases were matched with 35 controls each, 6 cases with 2 controls each, and 1 case with 3 controls) based on the number of available patients with similar characteristics found in the database. The matching was conducted according to the following criteria: age (range of 10 years from case reference, +/− 5 years), gender, predicted post-operative FEV1 (3 groups: <70%; 70–90%; >90%), and type of lobectomy performed. Data was collected and analyzed using Microsoft Excel software (version number 16.0) and Stata version 17 (StataCorp. 2021. Stata Statistical Software: Release 17. College Station, TX, USA: StataCorp LLC), respectively. A p-value < 0.05 was considered statistically significant.