https://orcid.org/0009-0005-4843-067X
Abstract
This work was designed to identify and summarize the current comparative data between digital pleural drainage devices and traditional analog pleural drainage devices in patients undergoing pulmonary resection. Outcomes of interest were hospital length of stay (LOS), chest tube duration, healthcare costs, incidence of prolonged air leak and patient and healthcare personnel satisfaction.
The PubMed and Web of Science databases were searched for randomized controlled trials (RCTs) and meta-analyses comparing outcomes between the use of digital and analog chest tube drainage devices.
Fourteen RCTs and 4 meta-analyses were identified. Eight RCTs found no difference in LOS or chest tube duration while 6 RCTs found a decrease in LOS and chest tube duration with digital devices. All the meta-analyses showed decreased LOS and chest tube duration with digital devices. Six out of 8 studies that compared the incidence of prolonged air leak showed no difference between digital and analog cohorts. Two RCTs and 1 meta-analysis reported decreased costs with digital systems. Two out of 3 RCTs reported improved patient and healthcare personnel satisfaction with digital systems while 1 RCT reported no difference.
While the data comparing length of stay and chest tube duration are mixed, the introduction of digital pleural drainage devices has provided a more objective assessment of air leak flow rates and potentially a more accurate assessment of the timing of air leak resolution after pulmonary resection. Further studies are needed to better delineate the cost utility and ideal scenarios for the use of digital devices.
INTRODUCTION
Background
Air leaks after elective pulmonary resection continue to be a significant burden for the patient and for the healthcare system impacting patient satisfaction, pain management, hospital length of stay (LOS), and ultimately healthcare costs. Efforts have been made to mitigate postoperative air leaks with changes in intraoperative techniques as well as changes in postoperative chest tube approach. Unfortunately, institutional customs may sometimes play a role in chest tube management strategies when objective clinical data are available to support a more streamlined, evidence-based management. The Society of Thoracic Surgeons has recently taken an initiative to provide consensus guidelines for postoperative chest tube management. Topics encompassed in the work groups review included number of chest tubes, chest tube size, suction versus water seal, volume threshold for chest tube removal, utilization of postoperative imaging, management of prolonged air leaks (PAL), and consensus recommendations on the use of traditional air leak monitoring devices versus digital air leak monitoring systems.
Historically, the timing of chest tube removal has varied depending on the threshold for chest tube removal utilized by the surgical team and can be inclusive of air leak criteria as well as volume of chest tube effluent [1–4]. Utilization of analogue chest drainage systems to assess for air leaks can be inconsistent secondary to physician variability in the frequency of assessment and interpretation of air leak resolution. More recently, digital drainage systems have been introduced in order to minimize this interobserver variability and to mitigate the potential delayed decision-making on chest tube removal that can also increase LOS. Digital chest drainage systems, which provide objective data on the rate of air leak and a more precise confirmation of air leak cessation, may improve efficiency in managing chest tubes postoperatively. Secondary benefits include a built-in suction mechanism that provides a more physiologic application of negative intrathoracic pressure and consistent tracking of fluid drainage.
Digital chest tube management devices
Recently, digital chest drainage systems have been introduced and these devices have the benefit of providing a fixed intrapleural pressure as well as containing sensors that detect pleural pressure and air flow [5, 6]. In the presence of an air leak, these devices will exert an appropriate amount of suction to maintain a fixed pre-set negative intrapleural pressure. The first digital air flow monitoring device for evaluating air leaks was the AIRFIX (TEUP's Ltd, Deutschlandsberg, Austria). This device was attached in line with the chest tube with an inflow port connected to the chest tube and the outflow port connected to the drainage system and was not a self-contained chest drainage system with built-in suction like the digital systems in use now [7]. There have been several digital pleural drainage systems available including DigiVent Chest Drainage System (Digivent™, Millicore AB, Danderyd, Sweden), Thopaz (Medela Baar, Switzerland), Drentech Palm Evo (Redax, Italy), Thoraguard (Centese, Omaha, Nebraska), and ATMOS (MedizinTechnik, Sulz, Germany). DigiVent, although no longer commercially available, was the first available self-contained system and the results of an efficacy trial were published in 2007 [6]. The DigiVent system had 2 sensors—1 for flow and 1 for pressure—and could record instances of air leak so that the average rate of flow over a given time period could be determined. The Thopaz system also senses fluid output rate in addition to air flow and pressure [8]. The Drentech system is a portable vacuum unit that can be attached to a standard disposable drainage system and has a display that shows current air leak rate, mean air leak over the last hour and real-time minimum and maximum intrapleural pressures. Data from this device can be transferred to a USB drive [9]. The Thoraguard system indirectly measures air leak using a tachometer to measure the work of the pump to maintain a constant intrapleural pressure, which allows for continuous monitoring of air leak with a longitudinal graphical display of data, intrapleural pressure, and fluid drainage. The device also has a built-in automated clog clearance system [10]. Several of these digital systems, including the Thoraguard, Thopaz, and Atmos systems also have alarms to alert staff of issues such as obstruction, catheter disconnection, clogging in the canister, excessive air leak, low battery and inability to reach the target intrapleural pressure [11–13].
The potential advantages of digital chest drainage systems include reduced interobserver variability in assessment of air leak, real-time monitoring of air leaks rates and intrapleural pressure further optimizing the timing of chest tube removal. Many of these devices are able to sense intra-pleural pressure and adjust the suction pressure to physiologic parameters, ideally preventing PAL. The primary practical objective of these features is to decrease chest tube duration and consequently decrease hospital LOS. The potential obstacle to routine utilization of the digital devices is the increased device-related cost relative to the analogue systems which could conceivably be overcome with more expeditious chest tube removal resulting in decreased hospital LOS. The objective of this narrative review was to identify and examine the existing literature comparing outcomes with the use of digital versus analogue chest drainage systems.
METHODS
The PubMed database was searched using the terms ‘digital chest drain’ or ‘digital chest tube’. The Web of Science database was searched using the query TS = (conventional NEAR/5 drain*) OR TS = (traditional NEAR/5 drain*) OR TS = (analogue* NEAR/5 drain*) AND TS = (digital NEAR/5 drain*) OR TS = (electronic NEAR/5 drain*). The initial search yielded retrospective studies, randomized controlled trials (RCTs), meta-analyses, systematic reviews, observational studies, and propensity matched studies. Only studies involving adult subjects undergoing pulmonary resection comparing outcomes with the use of digital versus analogue chest drainage systems were included in this review. Studies pertaining to management of chest tubes after cardiac surgery, for treatment of spontaneous pneumothorax, or for treatment of pleural effusion were excluded (Fig. 1).
Flowchart of literature search strategy. RCT: randomized controlled trial.
RESULTS
Chest tube duration and hospital length of stay
There were 14 RCTs that compared duration of chest drainage and LOS with digital versus analogue systems—11 of which were single-center studies while 3 were multicentre studies (Table 1) [14–27]. These trials ranged in sample size from 31 to 465 and included patients undergoing pulmonary resection for malignant or benign diseases via thoracotomy or a mix of thoracotomy and thoracoscopy. Eight of these RCTs showed no significant difference in chest tube duration or LOS. A single-center study by Takamochi et al. that included 299 patients found no difference in chest tube duration or LOS between the cohorts (3 vs 2 days, P = 0.149; 7 vs 6 days, P = 0.548) [19]. Interestingly, more patients in the analogue group had air leaks recorded postoperatively. A subgroup analysis of the patients with air leaks showed no difference in chest tube duration, LOS, duration of air leak, or PAL between the groups. A study by Gilbert et al. [16] demonstrated similar chest tube duration (air leak absent: 3 vs 2.9 days, P = 0.05; air leak present: 5.6 vs 4.9 days, P = 0.11) and LOS (air leak absent: 4 vs 4 days, P = 0.09; air leak present: 6 vs 6 days, P = 0.36) between cohorts regardless of whether an air leak was present or absent early postoperatively.
Randomized controlled trials comparing outcomes in digital versus analogue systems
| Study | Year | Country | Location type | Number of patients | Type of resection (% lobectomy, %segmentectomy or wedge) | Approach (% Thoracotomy, % VATS) | Device | Chest tube management | Chest tube removal protocol | Air leak duration | Chest tube duration (analog versus digital) | Hospital LOS (analog versus digital) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Bertolaccini et al. [14] | 2011 | Italy | Single centre | 98 | Pulmonary resection excluding pneumonectomy (47%, 53%) | Not specified | Drentech Palm | 24 Fr spiral drain and 28 Fr standard chest tube. Connected to −15 cmH2O suction until POD1 X-ray | Analog: No air leak, drainage <250 cc/day, no abnormal findings on X-ray Digital: No air leak for 6 h | Not reported | 3.5 vs 3.1 days (first chest tube, mean), P = 0.06 6.1 vs 5.5 days (second chest tube), P = 0.06 | 7.1 vs 6.5 days (mean), P = 0.09 |
| Chiappetta et al. [15] | 2018 | Italy | Single center | 95 | Lobectomy, segmentectomy, wedge resection (53%, 47%) | Thoracotomy (100%, 0%) | Drentech Simple Plus PALM | Single 28 Fr chest tube placed after wedge resection, chest tubes (28 and 24 Fr) placed after major anatomic resection. No suction applied unless incomplete expansion on X-ray | Analog: No clinical signs of air leak (bubbles, subcutaneous emphysema), drainage <250 cc/day Digital: Air leak <0.5 l/h for 12 h, drainage <250/day | Not reported | 4.6 vs 4.1 days (mean), P = 0.4 | 6.2 vs 5.8 days (mean), P = 0.4 |
| Gilbert et al. [16] | 2015 | Canada | Single center | 172 | Sublobar or lobar pulmonary resection (77%, 23%) | VATS (0%, 100%) | Thopaz | Suction applied immediately after surgery and stopped POD1 if chest X-ray with ≤30% pneumothorax and no significant subcutaneous emphysema | Analog: No bubbling, drainage ≤250 cc/day Digital: Air leak ≤40 cc/min, on negative pressure (>8 mmHg) or ≤20 cc/min, on gravity (≤8 mmHg) for a minimum of 12 h, drainage ≤250 cc/day | 4.7 vs 3.7 days, P = 0.14 | Air leak absent: 3 vs 2.9 days (median), P = 0.05 Air leak present: 5.6 vs 4.9 days (median), P = 0.11 | Air leak absent: 4 vs 4 days (median), P = 0.09 Air leak present: 6 vs 6 days (median), P = 0.36 |
| Marjanski et al. [17] | 2013 | Poland | Single center | 64 | Lobectomy (100%, 0%) | Not specified | Thopaz | Single 28 Fr chest tube, suction at −15 cmH2O for 2 days, then gravity | Analog: No bubbling with cough or Valsalva, drainage <350 cc/day Digital: Air leak 0–20 cc/min over 6 h | Not reported | 4 vs 4 days (median), P value not reported | 5 vs 6 days (median), P value not reported |
| Plourde et al. [18] | 2018 | Canada | Single center | 215 | Lobectomy or segmentectomy (96%, 4%) | Thoracotomy or VATS (16%, 84%) | ATMOS | Single 28 Fr chest tube placed to −20 cmH2O suction. POD1 analog group placed to water seal, digital placed to −5 cmH2O. −20 cm suction applied if increasing pneumothorax, symptomatic pneumothorax, or increasing subcutaneous emphysema | Analog: No bubbling after 3 sequential coughs Digital: air flow <0.03 l/min for 12 h | Not reported | 3 vs 3 days (median), P = 0.20 | 5 vs 4 days (median), P = 0.47 |
| Takamochi et al. [19] | 2018 | Japan | Single center | 299 | Segmentectomy or lobectomy (79%, 21%) | Thoracotomy (100%, 0%) | Thopaz | Analog: −10 cmH2O until POD1 then water seal. Chest tube replaced to −10 cmH2O when checking for air leaks Digital: −10 cmH2O | Analog: No bubbling, drainage <300 cc/day. Chest tube clamped 24 h if air leak suspected Digital: Air leak rate <20 cc/min for >12 h and drainage ≤300 cc/day. | Not reported | 3 vs 2 days (median), P = 0.149 | 7 vs 6 days (median), P = 0.548 |
| Lijkendijk et al. [20] | 2015 | Denmark | Single center | 105 | Lobectomy (100%, 0%) | Thoracotomy or VATS (61%, 39%) | Thopaz | Analog: Placed to gravity immediately Digital: −15 cmH2O suction | Analog: No bubbling in water seal chamber with cough, <200 cc drainage or <400 cc on POD1 Digital: A leak ≤20 cc/min for 6 h or ≤50 cc/min for 12 h, <200 cc drainage on POD0 or <400 cc on POD1 | Not reported | 46.5 vs 41 h (median), P = 0.397 | 5 vs 4 days (median), P = 0.137 |
| Brunelli et al. [21] | 2010 | Italy | Single center | 159 | Lobectomy (100%, 0%) | Thoracotomy (100%, 0%) | Digivent | Two 28 Fr chest tub. −15 cmH2O suction on POD0. Alternating suction with −15 cmH2O at night and no suction during day in POD1 and after. | Both groups: Posterior chest tube removed when drainage <400 cc/day and no air leak Digital: Air leak 0 cc/min over 6 h, chest X-ray with adequate re-expansion, drainage <400 cc/day. If air leak <15 cc/min over 6 h and adequate re-expansion on chest X-ray, 12–24 h clamp trial. Analog: No bubbling, chest X-ray with adequate re-expansion, drainage <400 cc/day. 12 h clamp trial if equivocal. | Not reported | 4.9 vs 4 days (mean), P = 0.0007 | 6.3 vs 5.4 days (mean), P = 0.007 |
| Cerfolio and Bryant [22] | 2008 | USA | Single center | 100 | Elective pulmonary resection (55%, 45%) | Thoracotomy (100%, 0%) | Digivent | −20 cm suction on POD0 then water seal POD1. Daily chest X-ray. Placed back to suction if increasing pneumothorax or subcutaneous emphysema Analog: If air leak present, classified by Cerfolio classification system then placed on digital system for 2 h to quantify leak in cm3/min before placing back on analogue system Digital: If air leak present, recorded in cm3/min then placed on analogue system for 2 h to classify leak by Cerfolio system before placing back on digital system | Drainage <450 cm3/day, no air leak on either system | Not reported | 3.9 vs 3.1 days (mean), P = 0.034 | 4 vs 3.3 days (mean), P = 0.055 |
| De Waele [23] | 2017 | Canada | Single center | 103 | Anatomic segmentectomy, lobectomy or bilobectomy (68%, 32%) | Thoracotomy or VATS (53%, 47%) | Thopaz | 2 28-Fr chest tubes Analog: −20 cmH2O suction Digital: Intermittent suction to maintain −20 cmH2O pleural pressure | Analog: <Grade 2 on Cerfolio scale, <350 cc/day or <150 cc for 8 h, lung re-expanded on chest X-ray Digital: <40 cc/min over 8 h, <350 cc/day or <150 cc for 8 h, lung re-expanded on chest X-ray | Not reported | 2.5 vs 2.3 days (mean), P = 0.055 | 4.9 vs 4.8 days (mean), P = 0.403 |
| Filosso et al. [24] | 2010 | Italy | Single center | 31 | Lobectomy (100%, 0%) | Thoracotomy (100%, 0%) | Drentech Palm | 1 chest tube 24–28 Fr, −20 cmH2O suction | Analog: POD2 or later, No detectable air leak, <250 cc/day Digital: POD2 or later, <20 cc/min over 6 h, <250 cc/day | Not reported | Chest tube duration decreased with digital (actual mean/median not reported), P = 0.001 (first chest tube), P = 0.005 (second chest tube) | LOS decreased with digital (actual mean/median not reported), P = 0.00001 |
| Pompili et al. [25] | 2014 | Italy, USA, UK, China | Multicenter | 381 | Pulmonary lobectomy or segmentectomy (84%, 16%) | Thoracotomy or VATS (81%, 19%) | Thopaz | –20 cmH2O on POD0 Analog: No suction on POD1 Digital: −8 cmH2O on POD1 | Analog: No detectable air leak, <300–400 cc/day drainage Digital: <30 cc/min for 8 h, <300–400 cc/day drainage | 1.0 vs 2.2 days, P = 0.001 | 4.7 vs 3.6 days (mean), P = 0.0001 | 5.6 vs 4.6 days (mean) P < 0.0001 |
| Comacchio et al. [26] | 2023 | Italy | Multicenter | 465 | Lobectomy (100%, 0%) | VATS (0%, 100%) | Drentech Palm Evo | One 28 Fr chest tube, −20 cmH2O suction for 24 h then no suction | Analog: No bubbling in water seal chamber with cough, <300 cc/day drainage, re-expansion on chest X-ray Digital: Air leak <20 cc/min for 8 h, <300 cc/day drainage, re-expansion on chest X-ray | 0 vs 0 days, P = 0.92 | 4 vs 3 days (median), P = 0.001 | 5 vs 4 days (median), P = 0.035 |
| Mendogni et al. [27] | 2021 | Italy | Multicenter | 209 | Lobectomy (100%, 0%) | VATS (0%, 100%) | Drentech Palm Evo | One 28 Fr chest tube | Analog: No detectable air leak, re-expansion on chest X-ray, <300 cc/day drainage Digital: Air leak <20 cc/min for 8 h, <300 cc/day drainage | Not reported | 3.8 vs 1.4 days (median), P = 0.999 | Not reported |
| Study | Year | Country | Location type | Number of patients | Type of resection (% lobectomy, %segmentectomy or wedge) | Approach (% Thoracotomy, % VATS) | Device | Chest tube management | Chest tube removal protocol | Air leak duration | Chest tube duration (analog versus digital) | Hospital LOS (analog versus digital) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Bertolaccini et al. [14] | 2011 | Italy | Single centre | 98 | Pulmonary resection excluding pneumonectomy (47%, 53%) | Not specified | Drentech Palm | 24 Fr spiral drain and 28 Fr standard chest tube. Connected to −15 cmH2O suction until POD1 X-ray | Analog: No air leak, drainage <250 cc/day, no abnormal findings on X-ray Digital: No air leak for 6 h | Not reported | 3.5 vs 3.1 days (first chest tube, mean), P = 0.06 6.1 vs 5.5 days (second chest tube), P = 0.06 | 7.1 vs 6.5 days (mean), P = 0.09 |
| Chiappetta et al. [15] | 2018 | Italy | Single center | 95 | Lobectomy, segmentectomy, wedge resection (53%, 47%) | Thoracotomy (100%, 0%) | Drentech Simple Plus PALM | Single 28 Fr chest tube placed after wedge resection, chest tubes (28 and 24 Fr) placed after major anatomic resection. No suction applied unless incomplete expansion on X-ray | Analog: No clinical signs of air leak (bubbles, subcutaneous emphysema), drainage <250 cc/day Digital: Air leak <0.5 l/h for 12 h, drainage <250/day | Not reported | 4.6 vs 4.1 days (mean), P = 0.4 | 6.2 vs 5.8 days (mean), P = 0.4 |
| Gilbert et al. [16] | 2015 | Canada | Single center | 172 | Sublobar or lobar pulmonary resection (77%, 23%) | VATS (0%, 100%) | Thopaz | Suction applied immediately after surgery and stopped POD1 if chest X-ray with ≤30% pneumothorax and no significant subcutaneous emphysema | Analog: No bubbling, drainage ≤250 cc/day Digital: Air leak ≤40 cc/min, on negative pressure (>8 mmHg) or ≤20 cc/min, on gravity (≤8 mmHg) for a minimum of 12 h, drainage ≤250 cc/day | 4.7 vs 3.7 days, P = 0.14 | Air leak absent: 3 vs 2.9 days (median), P = 0.05 Air leak present: 5.6 vs 4.9 days (median), P = 0.11 | Air leak absent: 4 vs 4 days (median), P = 0.09 Air leak present: 6 vs 6 days (median), P = 0.36 |
| Marjanski et al. [17] | 2013 | Poland | Single center | 64 | Lobectomy (100%, 0%) | Not specified | Thopaz | Single 28 Fr chest tube, suction at −15 cmH2O for 2 days, then gravity | Analog: No bubbling with cough or Valsalva, drainage <350 cc/day Digital: Air leak 0–20 cc/min over 6 h | Not reported | 4 vs 4 days (median), P value not reported | 5 vs 6 days (median), P value not reported |
| Plourde et al. [18] | 2018 | Canada | Single center | 215 | Lobectomy or segmentectomy (96%, 4%) | Thoracotomy or VATS (16%, 84%) | ATMOS | Single 28 Fr chest tube placed to −20 cmH2O suction. POD1 analog group placed to water seal, digital placed to −5 cmH2O. −20 cm suction applied if increasing pneumothorax, symptomatic pneumothorax, or increasing subcutaneous emphysema | Analog: No bubbling after 3 sequential coughs Digital: air flow <0.03 l/min for 12 h | Not reported | 3 vs 3 days (median), P = 0.20 | 5 vs 4 days (median), P = 0.47 |
| Takamochi et al. [19] | 2018 | Japan | Single center | 299 | Segmentectomy or lobectomy (79%, 21%) | Thoracotomy (100%, 0%) | Thopaz | Analog: −10 cmH2O until POD1 then water seal. Chest tube replaced to −10 cmH2O when checking for air leaks Digital: −10 cmH2O | Analog: No bubbling, drainage <300 cc/day. Chest tube clamped 24 h if air leak suspected Digital: Air leak rate <20 cc/min for >12 h and drainage ≤300 cc/day. | Not reported | 3 vs 2 days (median), P = 0.149 | 7 vs 6 days (median), P = 0.548 |
| Lijkendijk et al. [20] | 2015 | Denmark | Single center | 105 | Lobectomy (100%, 0%) | Thoracotomy or VATS (61%, 39%) | Thopaz | Analog: Placed to gravity immediately Digital: −15 cmH2O suction | Analog: No bubbling in water seal chamber with cough, <200 cc drainage or <400 cc on POD1 Digital: A leak ≤20 cc/min for 6 h or ≤50 cc/min for 12 h, <200 cc drainage on POD0 or <400 cc on POD1 | Not reported | 46.5 vs 41 h (median), P = 0.397 | 5 vs 4 days (median), P = 0.137 |
| Brunelli et al. [21] | 2010 | Italy | Single center | 159 | Lobectomy (100%, 0%) | Thoracotomy (100%, 0%) | Digivent | Two 28 Fr chest tub. −15 cmH2O suction on POD0. Alternating suction with −15 cmH2O at night and no suction during day in POD1 and after. | Both groups: Posterior chest tube removed when drainage <400 cc/day and no air leak Digital: Air leak 0 cc/min over 6 h, chest X-ray with adequate re-expansion, drainage <400 cc/day. If air leak <15 cc/min over 6 h and adequate re-expansion on chest X-ray, 12–24 h clamp trial. Analog: No bubbling, chest X-ray with adequate re-expansion, drainage <400 cc/day. 12 h clamp trial if equivocal. | Not reported | 4.9 vs 4 days (mean), P = 0.0007 | 6.3 vs 5.4 days (mean), P = 0.007 |
| Cerfolio and Bryant [22] | 2008 | USA | Single center | 100 | Elective pulmonary resection (55%, 45%) | Thoracotomy (100%, 0%) | Digivent | −20 cm suction on POD0 then water seal POD1. Daily chest X-ray. Placed back to suction if increasing pneumothorax or subcutaneous emphysema Analog: If air leak present, classified by Cerfolio classification system then placed on digital system for 2 h to quantify leak in cm3/min before placing back on analogue system Digital: If air leak present, recorded in cm3/min then placed on analogue system for 2 h to classify leak by Cerfolio system before placing back on digital system | Drainage <450 cm3/day, no air leak on either system | Not reported | 3.9 vs 3.1 days (mean), P = 0.034 | 4 vs 3.3 days (mean), P = 0.055 |
| De Waele [23] | 2017 | Canada | Single center | 103 | Anatomic segmentectomy, lobectomy or bilobectomy (68%, 32%) | Thoracotomy or VATS (53%, 47%) | Thopaz | 2 28-Fr chest tubes Analog: −20 cmH2O suction Digital: Intermittent suction to maintain −20 cmH2O pleural pressure | Analog: <Grade 2 on Cerfolio scale, <350 cc/day or <150 cc for 8 h, lung re-expanded on chest X-ray Digital: <40 cc/min over 8 h, <350 cc/day or <150 cc for 8 h, lung re-expanded on chest X-ray | Not reported | 2.5 vs 2.3 days (mean), P = 0.055 | 4.9 vs 4.8 days (mean), P = 0.403 |
| Filosso et al. [24] | 2010 | Italy | Single center | 31 | Lobectomy (100%, 0%) | Thoracotomy (100%, 0%) | Drentech Palm | 1 chest tube 24–28 Fr, −20 cmH2O suction | Analog: POD2 or later, No detectable air leak, <250 cc/day Digital: POD2 or later, <20 cc/min over 6 h, <250 cc/day | Not reported | Chest tube duration decreased with digital (actual mean/median not reported), P = 0.001 (first chest tube), P = 0.005 (second chest tube) | LOS decreased with digital (actual mean/median not reported), P = 0.00001 |
| Pompili et al. [25] | 2014 | Italy, USA, UK, China | Multicenter | 381 | Pulmonary lobectomy or segmentectomy (84%, 16%) | Thoracotomy or VATS (81%, 19%) | Thopaz | –20 cmH2O on POD0 Analog: No suction on POD1 Digital: −8 cmH2O on POD1 | Analog: No detectable air leak, <300–400 cc/day drainage Digital: <30 cc/min for 8 h, <300–400 cc/day drainage | 1.0 vs 2.2 days, P = 0.001 | 4.7 vs 3.6 days (mean), P = 0.0001 | 5.6 vs 4.6 days (mean) P < 0.0001 |
| Comacchio et al. [26] | 2023 | Italy | Multicenter | 465 | Lobectomy (100%, 0%) | VATS (0%, 100%) | Drentech Palm Evo | One 28 Fr chest tube, −20 cmH2O suction for 24 h then no suction | Analog: No bubbling in water seal chamber with cough, <300 cc/day drainage, re-expansion on chest X-ray Digital: Air leak <20 cc/min for 8 h, <300 cc/day drainage, re-expansion on chest X-ray | 0 vs 0 days, P = 0.92 | 4 vs 3 days (median), P = 0.001 | 5 vs 4 days (median), P = 0.035 |
| Mendogni et al. [27] | 2021 | Italy | Multicenter | 209 | Lobectomy (100%, 0%) | VATS (0%, 100%) | Drentech Palm Evo | One 28 Fr chest tube | Analog: No detectable air leak, re-expansion on chest X-ray, <300 cc/day drainage Digital: Air leak <20 cc/min for 8 h, <300 cc/day drainage | Not reported | 3.8 vs 1.4 days (median), P = 0.999 | Not reported |
LOS: length of stay; POD: post-operative day; VATS: video-assisted thoracoscopic surgery.
Randomized controlled trials comparing outcomes in digital versus analogue systems
| Study | Year | Country | Location type | Number of patients | Type of resection (% lobectomy, %segmentectomy or wedge) | Approach (% Thoracotomy, % VATS) | Device | Chest tube management | Chest tube removal protocol | Air leak duration | Chest tube duration (analog versus digital) | Hospital LOS (analog versus digital) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Bertolaccini et al. [14] | 2011 | Italy | Single centre | 98 | Pulmonary resection excluding pneumonectomy (47%, 53%) | Not specified | Drentech Palm | 24 Fr spiral drain and 28 Fr standard chest tube. Connected to −15 cmH2O suction until POD1 X-ray | Analog: No air leak, drainage <250 cc/day, no abnormal findings on X-ray Digital: No air leak for 6 h | Not reported | 3.5 vs 3.1 days (first chest tube, mean), P = 0.06 6.1 vs 5.5 days (second chest tube), P = 0.06 | 7.1 vs 6.5 days (mean), P = 0.09 |
| Chiappetta et al. [15] | 2018 | Italy | Single center | 95 | Lobectomy, segmentectomy, wedge resection (53%, 47%) | Thoracotomy (100%, 0%) | Drentech Simple Plus PALM | Single 28 Fr chest tube placed after wedge resection, chest tubes (28 and 24 Fr) placed after major anatomic resection. No suction applied unless incomplete expansion on X-ray | Analog: No clinical signs of air leak (bubbles, subcutaneous emphysema), drainage <250 cc/day Digital: Air leak <0.5 l/h for 12 h, drainage <250/day | Not reported | 4.6 vs 4.1 days (mean), P = 0.4 | 6.2 vs 5.8 days (mean), P = 0.4 |
| Gilbert et al. [16] | 2015 | Canada | Single center | 172 | Sublobar or lobar pulmonary resection (77%, 23%) | VATS (0%, 100%) | Thopaz | Suction applied immediately after surgery and stopped POD1 if chest X-ray with ≤30% pneumothorax and no significant subcutaneous emphysema | Analog: No bubbling, drainage ≤250 cc/day Digital: Air leak ≤40 cc/min, on negative pressure (>8 mmHg) or ≤20 cc/min, on gravity (≤8 mmHg) for a minimum of 12 h, drainage ≤250 cc/day | 4.7 vs 3.7 days, P = 0.14 | Air leak absent: 3 vs 2.9 days (median), P = 0.05 Air leak present: 5.6 vs 4.9 days (median), P = 0.11 | Air leak absent: 4 vs 4 days (median), P = 0.09 Air leak present: 6 vs 6 days (median), P = 0.36 |
| Marjanski et al. [17] | 2013 | Poland | Single center | 64 | Lobectomy (100%, 0%) | Not specified | Thopaz | Single 28 Fr chest tube, suction at −15 cmH2O for 2 days, then gravity | Analog: No bubbling with cough or Valsalva, drainage <350 cc/day Digital: Air leak 0–20 cc/min over 6 h | Not reported | 4 vs 4 days (median), P value not reported | 5 vs 6 days (median), P value not reported |
| Plourde et al. [18] | 2018 | Canada | Single center | 215 | Lobectomy or segmentectomy (96%, 4%) | Thoracotomy or VATS (16%, 84%) | ATMOS | Single 28 Fr chest tube placed to −20 cmH2O suction. POD1 analog group placed to water seal, digital placed to −5 cmH2O. −20 cm suction applied if increasing pneumothorax, symptomatic pneumothorax, or increasing subcutaneous emphysema | Analog: No bubbling after 3 sequential coughs Digital: air flow <0.03 l/min for 12 h | Not reported | 3 vs 3 days (median), P = 0.20 | 5 vs 4 days (median), P = 0.47 |
| Takamochi et al. [19] | 2018 | Japan | Single center | 299 | Segmentectomy or lobectomy (79%, 21%) | Thoracotomy (100%, 0%) | Thopaz | Analog: −10 cmH2O until POD1 then water seal. Chest tube replaced to −10 cmH2O when checking for air leaks Digital: −10 cmH2O | Analog: No bubbling, drainage <300 cc/day. Chest tube clamped 24 h if air leak suspected Digital: Air leak rate <20 cc/min for >12 h and drainage ≤300 cc/day. | Not reported | 3 vs 2 days (median), P = 0.149 | 7 vs 6 days (median), P = 0.548 |
| Lijkendijk et al. [20] | 2015 | Denmark | Single center | 105 | Lobectomy (100%, 0%) | Thoracotomy or VATS (61%, 39%) | Thopaz | Analog: Placed to gravity immediately Digital: −15 cmH2O suction | Analog: No bubbling in water seal chamber with cough, <200 cc drainage or <400 cc on POD1 Digital: A leak ≤20 cc/min for 6 h or ≤50 cc/min for 12 h, <200 cc drainage on POD0 or <400 cc on POD1 | Not reported | 46.5 vs 41 h (median), P = 0.397 | 5 vs 4 days (median), P = 0.137 |
| Brunelli et al. [21] | 2010 | Italy | Single center | 159 | Lobectomy (100%, 0%) | Thoracotomy (100%, 0%) | Digivent | Two 28 Fr chest tub. −15 cmH2O suction on POD0. Alternating suction with −15 cmH2O at night and no suction during day in POD1 and after. | Both groups: Posterior chest tube removed when drainage <400 cc/day and no air leak Digital: Air leak 0 cc/min over 6 h, chest X-ray with adequate re-expansion, drainage <400 cc/day. If air leak <15 cc/min over 6 h and adequate re-expansion on chest X-ray, 12–24 h clamp trial. Analog: No bubbling, chest X-ray with adequate re-expansion, drainage <400 cc/day. 12 h clamp trial if equivocal. | Not reported | 4.9 vs 4 days (mean), P = 0.0007 | 6.3 vs 5.4 days (mean), P = 0.007 |
| Cerfolio and Bryant [22] | 2008 | USA | Single center | 100 | Elective pulmonary resection (55%, 45%) | Thoracotomy (100%, 0%) | Digivent | −20 cm suction on POD0 then water seal POD1. Daily chest X-ray. Placed back to suction if increasing pneumothorax or subcutaneous emphysema Analog: If air leak present, classified by Cerfolio classification system then placed on digital system for 2 h to quantify leak in cm3/min before placing back on analogue system Digital: If air leak present, recorded in cm3/min then placed on analogue system for 2 h to classify leak by Cerfolio system before placing back on digital system | Drainage <450 cm3/day, no air leak on either system | Not reported | 3.9 vs 3.1 days (mean), P = 0.034 | 4 vs 3.3 days (mean), P = 0.055 |
| De Waele [23] | 2017 | Canada | Single center | 103 | Anatomic segmentectomy, lobectomy or bilobectomy (68%, 32%) | Thoracotomy or VATS (53%, 47%) | Thopaz | 2 28-Fr chest tubes Analog: −20 cmH2O suction Digital: Intermittent suction to maintain −20 cmH2O pleural pressure | Analog: <Grade 2 on Cerfolio scale, <350 cc/day or <150 cc for 8 h, lung re-expanded on chest X-ray Digital: <40 cc/min over 8 h, <350 cc/day or <150 cc for 8 h, lung re-expanded on chest X-ray | Not reported | 2.5 vs 2.3 days (mean), P = 0.055 | 4.9 vs 4.8 days (mean), P = 0.403 |
| Filosso et al. [24] | 2010 | Italy | Single center | 31 | Lobectomy (100%, 0%) | Thoracotomy (100%, 0%) | Drentech Palm | 1 chest tube 24–28 Fr, −20 cmH2O suction | Analog: POD2 or later, No detectable air leak, <250 cc/day Digital: POD2 or later, <20 cc/min over 6 h, <250 cc/day | Not reported | Chest tube duration decreased with digital (actual mean/median not reported), P = 0.001 (first chest tube), P = 0.005 (second chest tube) | LOS decreased with digital (actual mean/median not reported), P = 0.00001 |
| Pompili et al. [25] | 2014 | Italy, USA, UK, China | Multicenter | 381 | Pulmonary lobectomy or segmentectomy (84%, 16%) | Thoracotomy or VATS (81%, 19%) | Thopaz | –20 cmH2O on POD0 Analog: No suction on POD1 Digital: −8 cmH2O on POD1 | Analog: No detectable air leak, <300–400 cc/day drainage Digital: <30 cc/min for 8 h, <300–400 cc/day drainage | 1.0 vs 2.2 days, P = 0.001 | 4.7 vs 3.6 days (mean), P = 0.0001 | 5.6 vs 4.6 days (mean) P < 0.0001 |
| Comacchio et al. [26] | 2023 | Italy | Multicenter | 465 | Lobectomy (100%, 0%) | VATS (0%, 100%) | Drentech Palm Evo | One 28 Fr chest tube, −20 cmH2O suction for 24 h then no suction | Analog: No bubbling in water seal chamber with cough, <300 cc/day drainage, re-expansion on chest X-ray Digital: Air leak <20 cc/min for 8 h, <300 cc/day drainage, re-expansion on chest X-ray | 0 vs 0 days, P = 0.92 | 4 vs 3 days (median), P = 0.001 | 5 vs 4 days (median), P = 0.035 |
| Mendogni et al. [27] | 2021 | Italy | Multicenter | 209 | Lobectomy (100%, 0%) | VATS (0%, 100%) | Drentech Palm Evo | One 28 Fr chest tube | Analog: No detectable air leak, re-expansion on chest X-ray, <300 cc/day drainage Digital: Air leak <20 cc/min for 8 h, <300 cc/day drainage | Not reported | 3.8 vs 1.4 days (median), P = 0.999 | Not reported |
| Study | Year | Country | Location type | Number of patients | Type of resection (% lobectomy, %segmentectomy or wedge) | Approach (% Thoracotomy, % VATS) | Device | Chest tube management | Chest tube removal protocol | Air leak duration | Chest tube duration (analog versus digital) | Hospital LOS (analog versus digital) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Bertolaccini et al. [14] | 2011 | Italy | Single centre | 98 | Pulmonary resection excluding pneumonectomy (47%, 53%) | Not specified | Drentech Palm | 24 Fr spiral drain and 28 Fr standard chest tube. Connected to −15 cmH2O suction until POD1 X-ray | Analog: No air leak, drainage <250 cc/day, no abnormal findings on X-ray Digital: No air leak for 6 h | Not reported | 3.5 vs 3.1 days (first chest tube, mean), P = 0.06 6.1 vs 5.5 days (second chest tube), P = 0.06 | 7.1 vs 6.5 days (mean), P = 0.09 |
| Chiappetta et al. [15] | 2018 | Italy | Single center | 95 | Lobectomy, segmentectomy, wedge resection (53%, 47%) | Thoracotomy (100%, 0%) | Drentech Simple Plus PALM | Single 28 Fr chest tube placed after wedge resection, chest tubes (28 and 24 Fr) placed after major anatomic resection. No suction applied unless incomplete expansion on X-ray | Analog: No clinical signs of air leak (bubbles, subcutaneous emphysema), drainage <250 cc/day Digital: Air leak <0.5 l/h for 12 h, drainage <250/day | Not reported | 4.6 vs 4.1 days (mean), P = 0.4 | 6.2 vs 5.8 days (mean), P = 0.4 |
| Gilbert et al. [16] | 2015 | Canada | Single center | 172 | Sublobar or lobar pulmonary resection (77%, 23%) | VATS (0%, 100%) | Thopaz | Suction applied immediately after surgery and stopped POD1 if chest X-ray with ≤30% pneumothorax and no significant subcutaneous emphysema | Analog: No bubbling, drainage ≤250 cc/day Digital: Air leak ≤40 cc/min, on negative pressure (>8 mmHg) or ≤20 cc/min, on gravity (≤8 mmHg) for a minimum of 12 h, drainage ≤250 cc/day | 4.7 vs 3.7 days, P = 0.14 | Air leak absent: 3 vs 2.9 days (median), P = 0.05 Air leak present: 5.6 vs 4.9 days (median), P = 0.11 | Air leak absent: 4 vs 4 days (median), P = 0.09 Air leak present: 6 vs 6 days (median), P = 0.36 |
| Marjanski et al. [17] | 2013 | Poland | Single center | 64 | Lobectomy (100%, 0%) | Not specified | Thopaz | Single 28 Fr chest tube, suction at −15 cmH2O for 2 days, then gravity | Analog: No bubbling with cough or Valsalva, drainage <350 cc/day Digital: Air leak 0–20 cc/min over 6 h | Not reported | 4 vs 4 days (median), P value not reported | 5 vs 6 days (median), P value not reported |
| Plourde et al. [18] | 2018 | Canada | Single center | 215 | Lobectomy or segmentectomy (96%, 4%) | Thoracotomy or VATS (16%, 84%) | ATMOS | Single 28 Fr chest tube placed to −20 cmH2O suction. POD1 analog group placed to water seal, digital placed to −5 cmH2O. −20 cm suction applied if increasing pneumothorax, symptomatic pneumothorax, or increasing subcutaneous emphysema | Analog: No bubbling after 3 sequential coughs Digital: air flow <0.03 l/min for 12 h | Not reported | 3 vs 3 days (median), P = 0.20 | 5 vs 4 days (median), P = 0.47 |
| Takamochi et al. [19] | 2018 | Japan | Single center | 299 | Segmentectomy or lobectomy (79%, 21%) | Thoracotomy (100%, 0%) | Thopaz | Analog: −10 cmH2O until POD1 then water seal. Chest tube replaced to −10 cmH2O when checking for air leaks Digital: −10 cmH2O | Analog: No bubbling, drainage <300 cc/day. Chest tube clamped 24 h if air leak suspected Digital: Air leak rate <20 cc/min for >12 h and drainage ≤300 cc/day. | Not reported | 3 vs 2 days (median), P = 0.149 | 7 vs 6 days (median), P = 0.548 |
| Lijkendijk et al. [20] | 2015 | Denmark | Single center | 105 | Lobectomy (100%, 0%) | Thoracotomy or VATS (61%, 39%) | Thopaz | Analog: Placed to gravity immediately Digital: −15 cmH2O suction | Analog: No bubbling in water seal chamber with cough, <200 cc drainage or <400 cc on POD1 Digital: A leak ≤20 cc/min for 6 h or ≤50 cc/min for 12 h, <200 cc drainage on POD0 or <400 cc on POD1 | Not reported | 46.5 vs 41 h (median), P = 0.397 | 5 vs 4 days (median), P = 0.137 |
| Brunelli et al. [21] | 2010 | Italy | Single center | 159 | Lobectomy (100%, 0%) | Thoracotomy (100%, 0%) | Digivent | Two 28 Fr chest tub. −15 cmH2O suction on POD0. Alternating suction with −15 cmH2O at night and no suction during day in POD1 and after. | Both groups: Posterior chest tube removed when drainage <400 cc/day and no air leak Digital: Air leak 0 cc/min over 6 h, chest X-ray with adequate re-expansion, drainage <400 cc/day. If air leak <15 cc/min over 6 h and adequate re-expansion on chest X-ray, 12–24 h clamp trial. Analog: No bubbling, chest X-ray with adequate re-expansion, drainage <400 cc/day. 12 h clamp trial if equivocal. | Not reported | 4.9 vs 4 days (mean), P = 0.0007 | 6.3 vs 5.4 days (mean), P = 0.007 |
| Cerfolio and Bryant [22] | 2008 | USA | Single center | 100 | Elective pulmonary resection (55%, 45%) | Thoracotomy (100%, 0%) | Digivent | −20 cm suction on POD0 then water seal POD1. Daily chest X-ray. Placed back to suction if increasing pneumothorax or subcutaneous emphysema Analog: If air leak present, classified by Cerfolio classification system then placed on digital system for 2 h to quantify leak in cm3/min before placing back on analogue system Digital: If air leak present, recorded in cm3/min then placed on analogue system for 2 h to classify leak by Cerfolio system before placing back on digital system | Drainage <450 cm3/day, no air leak on either system | Not reported | 3.9 vs 3.1 days (mean), P = 0.034 | 4 vs 3.3 days (mean), P = 0.055 |
| De Waele [23] | 2017 | Canada | Single center | 103 | Anatomic segmentectomy, lobectomy or bilobectomy (68%, 32%) | Thoracotomy or VATS (53%, 47%) | Thopaz | 2 28-Fr chest tubes Analog: −20 cmH2O suction Digital: Intermittent suction to maintain −20 cmH2O pleural pressure | Analog: <Grade 2 on Cerfolio scale, <350 cc/day or <150 cc for 8 h, lung re-expanded on chest X-ray Digital: <40 cc/min over 8 h, <350 cc/day or <150 cc for 8 h, lung re-expanded on chest X-ray | Not reported | 2.5 vs 2.3 days (mean), P = 0.055 | 4.9 vs 4.8 days (mean), P = 0.403 |
| Filosso et al. [24] | 2010 | Italy | Single center | 31 | Lobectomy (100%, 0%) | Thoracotomy (100%, 0%) | Drentech Palm | 1 chest tube 24–28 Fr, −20 cmH2O suction | Analog: POD2 or later, No detectable air leak, <250 cc/day Digital: POD2 or later, <20 cc/min over 6 h, <250 cc/day | Not reported | Chest tube duration decreased with digital (actual mean/median not reported), P = 0.001 (first chest tube), P = 0.005 (second chest tube) | LOS decreased with digital (actual mean/median not reported), P = 0.00001 |
| Pompili et al. [25] | 2014 | Italy, USA, UK, China | Multicenter | 381 | Pulmonary lobectomy or segmentectomy (84%, 16%) | Thoracotomy or VATS (81%, 19%) | Thopaz | –20 cmH2O on POD0 Analog: No suction on POD1 Digital: −8 cmH2O on POD1 | Analog: No detectable air leak, <300–400 cc/day drainage Digital: <30 cc/min for 8 h, <300–400 cc/day drainage | 1.0 vs 2.2 days, P = 0.001 | 4.7 vs 3.6 days (mean), P = 0.0001 | 5.6 vs 4.6 days (mean) P < 0.0001 |
| Comacchio et al. [26] | 2023 | Italy | Multicenter | 465 | Lobectomy (100%, 0%) | VATS (0%, 100%) | Drentech Palm Evo | One 28 Fr chest tube, −20 cmH2O suction for 24 h then no suction | Analog: No bubbling in water seal chamber with cough, <300 cc/day drainage, re-expansion on chest X-ray Digital: Air leak <20 cc/min for 8 h, <300 cc/day drainage, re-expansion on chest X-ray | 0 vs 0 days, P = 0.92 | 4 vs 3 days (median), P = 0.001 | 5 vs 4 days (median), P = 0.035 |
| Mendogni et al. [27] | 2021 | Italy | Multicenter | 209 | Lobectomy (100%, 0%) | VATS (0%, 100%) | Drentech Palm Evo | One 28 Fr chest tube | Analog: No detectable air leak, re-expansion on chest X-ray, <300 cc/day drainage Digital: Air leak <20 cc/min for 8 h, <300 cc/day drainage | Not reported | 3.8 vs 1.4 days (median), P = 0.999 | Not reported |
LOS: length of stay; POD: post-operative day; VATS: video-assisted thoracoscopic surgery.
In contrast, there have been 6 RCTs that reported improvement in clinical outcomes with the use of digital systems. Brunelli et al. reported decreases in chest tube duration f (4.9 vs 4 days, P = 0.0007) and LOS (6.3 vs 5.4 days, P = 0.007) with the digital devices [21]. The largest controlled trials have been 2 multicentre trials by Pompili and Comacchio et al. with both studies reporting decreases in chest tube duration and LOS in the digital cohort of over a day [25, 26].
There have been 4 meta-analyses and systematic reviews comparing digital to analogue chest drainage systems [28–31]. All of these meta-analyses reported a decrease in hospital stay and chest tube duration (Table 2). The largest of these by Chang et al. [29] included 21 studies comparing digital versus analogue or suction versus non-suction chest tube drainage systems after pulmonary resection or spontaneous pneumothorax with a total of 3399 patients and analysed the outcomes of LOS, chest tube duration, and incidence of PAL. This included 16 studies with a total of 2124 patients that reported chest tube duration and 15 studies with a total of 1870 patients that reported LOS. The authors found a mean difference in chest tube duration of −0.68 days (95% CI −1.32 to −0.04) and LOS of −1.40 days (95% CI −2.20 to −0.60) between the digital and analogue groups. Interestingly, this meta-analysis also showed a decrease in LOS in non-suction chest tube drainage systems compared to traditional suction drainage systems (−1.05, 95% CI −1.91 to −0.18) but did not compare non-suction to digital systems. A meta-analysis by Wang et al. [30] also reported decreased chest tube duration and LOS in the digital group but these results were based on only 2 studies.
Meta-analyses comparing outcomes in digital versus analogue systems
| Study | Year | Number of studies | Number of patients | Chest tube duration (analog versus digital) | Hospital LOS (analog versus digital) |
|---|---|---|---|---|---|
| Zhou et al. [28] | 2023 | 12 | 2000 | SMD = −0.34 days, 95% CI = −0.51 to −0.18, I2 = 68.7%, P < 0.001 | MD = −0.79 days, 95% CI = −1.24 to −0.34, I2 = 89.8%, P < 0.001 |
| Chang et al. [29] | 2022 | 21 | 3399 | MD = −0.68 days, 95% CI—1.32 to −0.04 | MD = −1.40 days, 95% CI −2.20 to −0.60 |
| Wang et al. [30] | 2019 | 8 | 1487 | SMD = − 0.35, 95% CI −0.60 to −0.09, P = 0.008 | SMD = − 0.35, 95% CI −0.61 to −0.09, P = 0.007 |
| Zhou et al. [31] | 2018 | 10 | 1268 | WMD −0.72 days; 95% CI −1.03 to −0.40, P < 0.001 | WMD −0.97 days; 95% CI −1.46 to −0.48, P < 0.001 |
| Study | Year | Number of studies | Number of patients | Chest tube duration (analog versus digital) | Hospital LOS (analog versus digital) |
|---|---|---|---|---|---|
| Zhou et al. [28] | 2023 | 12 | 2000 | SMD = −0.34 days, 95% CI = −0.51 to −0.18, I2 = 68.7%, P < 0.001 | MD = −0.79 days, 95% CI = −1.24 to −0.34, I2 = 89.8%, P < 0.001 |
| Chang et al. [29] | 2022 | 21 | 3399 | MD = −0.68 days, 95% CI—1.32 to −0.04 | MD = −1.40 days, 95% CI −2.20 to −0.60 |
| Wang et al. [30] | 2019 | 8 | 1487 | SMD = − 0.35, 95% CI −0.60 to −0.09, P = 0.008 | SMD = − 0.35, 95% CI −0.61 to −0.09, P = 0.007 |
| Zhou et al. [31] | 2018 | 10 | 1268 | WMD −0.72 days; 95% CI −1.03 to −0.40, P < 0.001 | WMD −0.97 days; 95% CI −1.46 to −0.48, P < 0.001 |
CI: confidence interval; LOS: length of stay; MD: mean difference; SMD: standard mean difference; WMD: weighted mean difference.
Meta-analyses comparing outcomes in digital versus analogue systems
| Study | Year | Number of studies | Number of patients | Chest tube duration (analog versus digital) | Hospital LOS (analog versus digital) |
|---|---|---|---|---|---|
| Zhou et al. [28] | 2023 | 12 | 2000 | SMD = −0.34 days, 95% CI = −0.51 to −0.18, I2 = 68.7%, P < 0.001 | MD = −0.79 days, 95% CI = −1.24 to −0.34, I2 = 89.8%, P < 0.001 |
| Chang et al. [29] | 2022 | 21 | 3399 | MD = −0.68 days, 95% CI—1.32 to −0.04 | MD = −1.40 days, 95% CI −2.20 to −0.60 |
| Wang et al. [30] | 2019 | 8 | 1487 | SMD = − 0.35, 95% CI −0.60 to −0.09, P = 0.008 | SMD = − 0.35, 95% CI −0.61 to −0.09, P = 0.007 |
| Zhou et al. [31] | 2018 | 10 | 1268 | WMD −0.72 days; 95% CI −1.03 to −0.40, P < 0.001 | WMD −0.97 days; 95% CI −1.46 to −0.48, P < 0.001 |
| Study | Year | Number of studies | Number of patients | Chest tube duration (analog versus digital) | Hospital LOS (analog versus digital) |
|---|---|---|---|---|---|
| Zhou et al. [28] | 2023 | 12 | 2000 | SMD = −0.34 days, 95% CI = −0.51 to −0.18, I2 = 68.7%, P < 0.001 | MD = −0.79 days, 95% CI = −1.24 to −0.34, I2 = 89.8%, P < 0.001 |
| Chang et al. [29] | 2022 | 21 | 3399 | MD = −0.68 days, 95% CI—1.32 to −0.04 | MD = −1.40 days, 95% CI −2.20 to −0.60 |
| Wang et al. [30] | 2019 | 8 | 1487 | SMD = − 0.35, 95% CI −0.60 to −0.09, P = 0.008 | SMD = − 0.35, 95% CI −0.61 to −0.09, P = 0.007 |
| Zhou et al. [31] | 2018 | 10 | 1268 | WMD −0.72 days; 95% CI −1.03 to −0.40, P < 0.001 | WMD −0.97 days; 95% CI −1.46 to −0.48, P < 0.001 |
CI: confidence interval; LOS: length of stay; MD: mean difference; SMD: standard mean difference; WMD: weighted mean difference.
The protocols for chest tube management in RCTs have had significant variability. In some studies, such as that by Chiappetta et al. [15], no suction was applied in either group. In other studies, the amount of suction was different between the digital and analogue groups [18–20, 25]. For example, in multiple studies, after typically a 24-h period during which the chest tubes in both the digital and analogue arms were placed at a set suction rate varying from −10 to −20 mmHg, the digital devices were placed to suction while the analogue devices were placed to water seal until removal [14, 15, 16, 21]. The criteria for chest tube removal were also not consistent between control and experimental arms in the majority of the studies. For example, in the study by Chiappetta et al., the chest tubes were removed in the analogue group if there was no air leak detected, i.e. bubbling in the chamber, but chest tubes in the digital group could be removed if air flow detected by the device was <0.5 l/h over 12 h [15]. In a study by Filosso et al., the chest tube was not removed in either group unless flow was 0 ml/min or there was no bubbling detected but the protocol for those in the digital group only required this rate over a period of 8 h while the analogue group required no leak detected over a period of 24 h [32]. Variability in chest tube removal protocols has been a recurring dilemma among those using digital drainage devices as there is currently no consensus on appropriate flow rates and peak values for removal.
Cost comparisons between traditional and digital devices
There are some notable studies in the literature evaluating cost of a digital drainage system. In a randomized single-centre analysis of lobectomy patients in 2009, Brunelli et al. demonstrated a relative cost difference of 2391 vs 2867 euros (P < 0.008) in favour of a digital system [21]. The improvement in LOS afforded by direct air leak analysis offset the additional capital costs associated with the device. In a meta-analysis of the available literature, Zhou et al. found 2 studies that directly evaluated cost of a digital system with a cost savings of 443 Euros (P = −0.004). This study did include the previously mentioned Brunelli trial as the major contributor of patients for analysis, and overall, the cost difference was a consequence of reduced LOS [31]. The most formal analysis of cost of a digital system comes from the National Institute for Health and Care Excellence (NICE) [8]. This organization is part of the UK’s National Health Service, and they provide formal technology appraisals and clinical recommendations for care across the health system. The NHS is legally obliged to fund and resource treatments recommended by NICE’s technology appraisals. In their formal cost-effectiveness analysis, NICE provided a positive recommendation for national use of a digital pleural drainage system due to cost savings associated with the reduction in tube duration and LOS. This equated to a base-cost savings of 111.34 pounds per patient when used in a postoperative setting.
Prolonged air leak
There were 8 RCTs that assessed differences in incidence of PAL, typically defined as an air leak lasting >5–7 days (Table 3) [17, 19, 21, 23, 24, 26, 27]. Six of these studies showed a similar proportion of PAL between the cohorts while only 2 reported a decreased incidence of PAL with digital systems. There were also 3 meta-analyses that assessed incidence of PAL, none of which reported any differences between the digital and analogue systems [28, 29, 31].
Incidence of prolonged air leak in digital versus analogue drainage systems
| Study | Study type | Number of patients | Incidence of prolonged air leak by %, RR, or OR (digital versus analog) |
|---|---|---|---|
| Marjanski et al. [17] | RCT, single centre | 64 | 6 vs 9, P = 0.641 |
| Takamochi et al. [19] | RCT, single centre | 299 | 8.1 vs 7.9, P = 0.944 |
| Brunelli et al. [21] | RCT, single centre | 159 | 3 vs 0.6, P = 0.2 |
| De Waele et al. [23] | RCT, single centre | 103 | 1.7 vs 8, P = 0.025 |
| Filosso et al. [24] | RCT, single centre | 31 | Actual number not reported but no statistically significant difference, P = 0.23 |
| Pompili et al. [25] | RCT, multicentre | 381 | 2.6 vs 5.6, P = 0.03 |
| Comacchio et al. [26] | RCT, multicentre | 465 | 6 vs 5, P = 0.47 |
| Mendogni et al. [27] | RCT, multicentre | 209 | 2.6 vs 1.1, P = 0.999 |
| Zhou et al. [28] | Meta-analysis | Not reported, 5 studies | RR = 0.74, 95% CI = 0.36–1.49, I2 = 41.2%, P = 0.147 |
| Chang et al. [29] | Meta-analysis | 2709 | OR = 0.76, 95% CI: 0.42–1.39, P = 0.78 |
| Zhou et al. [31] | Meta-analysis | Not reported, 3 studies | RR = 0.36, 95% CI 0.04–3.17, P = 0.36 |
| Study | Study type | Number of patients | Incidence of prolonged air leak by %, RR, or OR (digital versus analog) |
|---|---|---|---|
| Marjanski et al. [17] | RCT, single centre | 64 | 6 vs 9, P = 0.641 |
| Takamochi et al. [19] | RCT, single centre | 299 | 8.1 vs 7.9, P = 0.944 |
| Brunelli et al. [21] | RCT, single centre | 159 | 3 vs 0.6, P = 0.2 |
| De Waele et al. [23] | RCT, single centre | 103 | 1.7 vs 8, P = 0.025 |
| Filosso et al. [24] | RCT, single centre | 31 | Actual number not reported but no statistically significant difference, P = 0.23 |
| Pompili et al. [25] | RCT, multicentre | 381 | 2.6 vs 5.6, P = 0.03 |
| Comacchio et al. [26] | RCT, multicentre | 465 | 6 vs 5, P = 0.47 |
| Mendogni et al. [27] | RCT, multicentre | 209 | 2.6 vs 1.1, P = 0.999 |
| Zhou et al. [28] | Meta-analysis | Not reported, 5 studies | RR = 0.74, 95% CI = 0.36–1.49, I2 = 41.2%, P = 0.147 |
| Chang et al. [29] | Meta-analysis | 2709 | OR = 0.76, 95% CI: 0.42–1.39, P = 0.78 |
| Zhou et al. [31] | Meta-analysis | Not reported, 3 studies | RR = 0.36, 95% CI 0.04–3.17, P = 0.36 |
CI: confidence interval; OR: odds ratio; RCT: randomized controlled trial; RR: risk ratio.
Incidence of prolonged air leak in digital versus analogue drainage systems
| Study | Study type | Number of patients | Incidence of prolonged air leak by %, RR, or OR (digital versus analog) |
|---|---|---|---|
| Marjanski et al. [17] | RCT, single centre | 64 | 6 vs 9, P = 0.641 |
| Takamochi et al. [19] | RCT, single centre | 299 | 8.1 vs 7.9, P = 0.944 |
| Brunelli et al. [21] | RCT, single centre | 159 | 3 vs 0.6, P = 0.2 |
| De Waele et al. [23] | RCT, single centre | 103 | 1.7 vs 8, P = 0.025 |
| Filosso et al. [24] | RCT, single centre | 31 | Actual number not reported but no statistically significant difference, P = 0.23 |
| Pompili et al. [25] | RCT, multicentre | 381 | 2.6 vs 5.6, P = 0.03 |
| Comacchio et al. [26] | RCT, multicentre | 465 | 6 vs 5, P = 0.47 |
| Mendogni et al. [27] | RCT, multicentre | 209 | 2.6 vs 1.1, P = 0.999 |
| Zhou et al. [28] | Meta-analysis | Not reported, 5 studies | RR = 0.74, 95% CI = 0.36–1.49, I2 = 41.2%, P = 0.147 |
| Chang et al. [29] | Meta-analysis | 2709 | OR = 0.76, 95% CI: 0.42–1.39, P = 0.78 |
| Zhou et al. [31] | Meta-analysis | Not reported, 3 studies | RR = 0.36, 95% CI 0.04–3.17, P = 0.36 |
| Study | Study type | Number of patients | Incidence of prolonged air leak by %, RR, or OR (digital versus analog) |
|---|---|---|---|
| Marjanski et al. [17] | RCT, single centre | 64 | 6 vs 9, P = 0.641 |
| Takamochi et al. [19] | RCT, single centre | 299 | 8.1 vs 7.9, P = 0.944 |
| Brunelli et al. [21] | RCT, single centre | 159 | 3 vs 0.6, P = 0.2 |
| De Waele et al. [23] | RCT, single centre | 103 | 1.7 vs 8, P = 0.025 |
| Filosso et al. [24] | RCT, single centre | 31 | Actual number not reported but no statistically significant difference, P = 0.23 |
| Pompili et al. [25] | RCT, multicentre | 381 | 2.6 vs 5.6, P = 0.03 |
| Comacchio et al. [26] | RCT, multicentre | 465 | 6 vs 5, P = 0.47 |
| Mendogni et al. [27] | RCT, multicentre | 209 | 2.6 vs 1.1, P = 0.999 |
| Zhou et al. [28] | Meta-analysis | Not reported, 5 studies | RR = 0.74, 95% CI = 0.36–1.49, I2 = 41.2%, P = 0.147 |
| Chang et al. [29] | Meta-analysis | 2709 | OR = 0.76, 95% CI: 0.42–1.39, P = 0.78 |
| Zhou et al. [31] | Meta-analysis | Not reported, 3 studies | RR = 0.36, 95% CI 0.04–3.17, P = 0.36 |
CI: confidence interval; OR: odds ratio; RCT: randomized controlled trial; RR: risk ratio.
Patient and healthcare personnel satisfaction
Some studies have reported on satisfaction from patients or healthcare personnel. Three RCTs assessed patient satisfaction when using digital versus analogue drainage systems [14, 22, 25]. In the study by Pompili et al., patients in the digital cohort reported improved ability to get out of bed, convenience for personnel and other patients, and comfort going home with the device. Other questions, for which no significant difference in responses was found, included the ability to walk alone, portability, social comfort, comfort sleeping at night, and whether they would want to change the device to another device they had seen other patients using. Bertolaccini et al. compared patients receiving the Drentech PALM digital device to analogue and also reported improved ease of mobility and ease of setup by nursing staff with the digital devices. In contrast, a study by Cerfolio et al. reported significantly more nursing staff rating the analogue system easy to set up compared to the Digivent digital system and similar portability ratings from patients. Two RCTs reported a decrease in the average time for the clinician to rate an air leak with digital devices [14, 22].
DISCUSSION
Prevention of air leaks and better chest tube management after pulmonary resection are 2 factors that have the greatest impact on the postoperative course and experience for patients and for the healthcare system. Although chest tube management strategies had evolved prior to the utilization of digital air leak monitoring devices there remains physician-dependent variability in the subjective assessment of air leaks and chest tube removal criteria. The digital air leak monitoring systems have provided a truly novel and innovative mechanism to allow for a more objective assessment of this problem. Digital estimation of the size of an air leak and more precise approximation of the timing of air leak resolution has the potential to expedite chest tube removal with a resultant decreased LOS and decreased hospital-associated costs. While the data discussed in the aforementioned clinical trials provide mixed results regarding the relative impact of digital systems on the timing of chest tube removal and LOS it is apparent that the digital systems provide advantages compared to the analogue devices.
Even within the context of randomized trials, there remain potential confounding variables and bias that are capable of impacting data interpretation within these studies. The inability to blind physicians to the type of device used for air leak monitoring may result in an implicit bias that could impact the timing of chest tube removal and hospital discharge. While in a randomized setting this bias could affect both cohorts, the tendency is for this bias to favour the new technology. Other confounding factors include variability in institutional criteria for the timing of chest tube removal including delays in the timing from documentation of air leak resolution to chest tube removal as well as delays in time to discharge after chest tube removal. In the largest and most contemporary clinical trial, Comacchio et al. randomized 465 patients undergoing VATS lobectomy and demonstrated no difference in the median duration of air leak [0 days (interquartile range 0–2) vs 0 days (interquartile range 0–2), P = 0.92] but chest tube removal occurred 1 day earlier with the digital device (3 vs 4 days, P = 0.001). This trial offers the largest, most homogeneous population of patients and makes a compelling argument for the benefits of the digital devices. This trial also highlights the subjective nature of chest tube removal using analogue devices in that despite documenting resolution of air leaks at the same time as the digital cohort, surgeons using the analogue devices delayed chest tube removal suggesting a lack of confidence in the interpretation of the resolution of air leaks with the analogue systems. Interestingly these data suggest that earlier removal of chest tubes after documenting air leak resolution with the analogue systems may have been appropriate, but the digital systems seem to provide a level of confidence in this decision that may otherwise vary between surgeons or based on surgeon experience.
Within the digital device cohorts, there has been variability in the threshold considered for air leak cessation, as measured by the air flow rate, that was deemed acceptable for chest tube removal ranging from 20 to 40 ml/h. Furthermore, trials had established varying lengths of time that the patient should be at this threshold before removing the chest tube [14, 15, 17–21, 25–27, 32]. This is an interesting byproduct from the utilization of the digital systems in the recognition that it may be safe to remove chest tubes with some minimal residual air leak without significant negative consequences. In contrast, the analogue device cohorts were required to have no discernable bubble in the chamber prior to chest tube removal.
Within the recent large trial by Comacchio et al., utilization of the digital devices again resulted in a shorter chest tube duration (3 vs 4 days, P = 0.001) that translated into a shorter LOS (4 vs 5 days, P = 0.035) demonstrating a close temporal relationship between chest tube removal and discharge. However, several of the previous trials exhibited inconsistency in the timing of hospital discharge relative to the timing of chest tube removal. In the trial of 299 patients by Takamochi et al., reported outcomes were similar between the cohorts with chest tube duration of 2 vs 3 days (P = 0.149). However, despite early chest tube removal the LOS was delayed by several days in both cohorts (6 vs 7 days, P = 0.548). While hospital discharge can be affected by many factors this again demonstrates the extensive variability in criteria for discharge after chest tube removal that may be influenced by surgeon, institution, or healthcare system. This is a very important factor in the clinical management of these patients as the motivation for decreasing chest tube duration lies in improving patient comfort but also in decreasing hospital-related costs that are mostly influenced by LOS.
One of the more challenging complications after high-risk pulmonary resection is a PAL which is associated with a longer chest tube duration with subsequent increased LOS and hospital-related costs. PAL can also be associated with other postoperative complications such as pneumonia and increased pain after surgery [33]. While analogue devices allow the physician to define the set amount of suction, the digital devices can be set to maintain a consistent pleural pressure adapting to variations in the patient’s pulmonary mechanics and effort. While some trials have suggested a decrease in the rate of PAL with the digital devices most of the studies have demonstrated similar rates of PAL. Continued progress in the understanding of the relationship between PAL and intrapleural pressure with the digital devices may ultimately help improve management strategies in these patients. Even the potential for the digital devices to improve the ability to predict a PAL early in the postoperative period may assist in the decision-making process of intervention strategies such as earlier application of ambulatory chest tube management.
While some surgeons have been reluctant to transition to the digital systems, even sceptics agree that providing an objective real-time assessment of air leak resolution is a vast improvement over the current analogue systems. The concern is the relative cost of the digital systems and how these can be justified given the financial constraints of our current healthcare system. The direct capital cost of a digital system is only 1 variable. Surgical teams must include the impact on indirect costs and outcomes measures as well, which may negate any capital investment in a formal cost-effectiveness analysis. While the disposable costs for a digital system (drainage canister) is on par with a standard analogue drainage canister, the reusable digital system requires an upfront cost of $2000–$3000 USD per unit. Specific prices are often negotiated between the specified company and individual institution and the costs of the digital systems may also vary depending on the country. No additional disposables are required, as a standard chest tube at the discretion of the surgeon can be utilized with either digital or analogue systems.
While the relative cost of a digital system is a concern with today’s healthcare cost constraints, an additional challenge is selecting eligible patients that may benefit from real-time air leak volume reporting. Using contemporary ERAS protocols, most uncomplicated sublobar resections can have the chest tube removed on postoperative day 0–1 rendering the digital systems unnecessary. In the current era, some complex anatomic lobar or sublobar resections may be candidates for early chest tube removal. While data suggests that in this context the digital devices improve safety with early removal, patients with an analogue pleural drainage system may also be able to undergo fast tracking with early chest tube removal without event [34]. At our institution, only analogue devices are in use currently and it early chest tube removal within postoperative day 0–1 is generally the practice, albeit with some variation between providers. It is important to note however that expedited removal of chest tubes after anatomic resection has not become mainstream, but is becoming increasingly employed at select institutions. A digital system will still provide objective data that can improve safety and lower the likelihood of pneumothorax after chest tube removal. Whether it is cost-effective in this setting is not known, but unlikely as costly technology used in situations to improve rare complications will often not meet a willingness-to-pay threshold. It is likely that a more judicious application of the digital systems in patients with air leaks after day 1 or 2 and in patients with more complex anatomic resections at higher risk for air leak are more likely to sustain a benefit from digital systems. A more selective algorithmic approach may be more promising in adjudicating the cost utility for digital systems moving forward [35].
The benefits of the digital devices in terms of improved patient satisfaction as well as improved satisfaction by healthcare personnel cannot be overstated. The digital systems are self-contained and therefore do not require wall-mounted suction resulting in improved early mobility and overall comfort. Additionally, healthcare staff report higher satisfaction with the setup and maintenance of the devices and the interpretation of air leak messages and obstruction warnings on the devices [2, 12, 18, 20]. These are all important considerations in the evaluation of these devices in a clinical practice. From an academic perspective, the utilization of digital systems has undoubtedly provided valuable new data on the natural history of air leaks and the chronology of air leak resolution. Inevitably, these data will benefit all patients and physicians including those utilizing analogue systems in their decision-making process.
The advancement in the current digital air leak monitoring systems has considerably improved our ability to objectively assess the air leak flow rate and to define the timing of air leak resolution more precisely. The digital devices can decrease the time to chest tube removal and decrease postoperative LOS after pulmonary resection. Utilization of the digital devices should be considered in the clinical setting when evaluating new comprehensive evidence-based protocols for perioperative management of patients undergoing pulmonary resection. Continued study and innovation of the digital devices will inevitably result in refinement of the ideal criteria for utilization and for improvement in the defined clinical outcomes and cost utility.
FUNDING
This paper was published as part of a supplement financially supported by Medela.
Conflict of interest: Brian Mitzman is a Proctor/Consultant at Intuitive Surgical. The other authors report no conflicts of interest.
DATA AVAILABILITY
Data supporting the findings of this study can be made available upon request to the corresponding author.
ETHICAL STATEMENT
The study did not include any data from individual subjects but was a narrative review of already published manuscripts.
REFERENCES
ABBREVIATIONS
- LOS
Length of stay
- NICE
National Institute for Health and Care Excellence
- PAL
Prolonged air leaks
- RCTs
Randomized controlled trials
