Michael H. Baumann, MD, FCCP; Praful B. Patel, MD; Chris W. Roney, MD; and Marcy F. Petrini, PhD, FCCP

Transcription

1 Comparison of Function of Commercially Available Pleural Drainage Units and Catheters* Michael H. Baumann, MD, FCCP; Praful B. Patel, MD; Chris W. Roney, MD; and Marcy F. Petrini, PhD, FCCP Purpose: Flow rates and pressures generated by commercially available pleural drainage units (PDUs) and flow rates through available pleural drainage catheters (PDCs) are not known. This information may be important clinically depending on the volume of air leak associated with a bronchopleural fistula. Design: Eight PDUs were assessed for flow rates at various suction levels and for the percent accuracy of suction pressures generated at various settings. Eleven commonly used PDCs were assessed for flow rates at various suction control levels. All devices were donated by their manufacturer. Flow rates and pressures were measured by a RT 200 Calibration Analyzer (Timeter Instrument Corporation; St. Louis, MO) at body temperature, ambient pressure, saturated with water vapor. Five devices of each type were tested. Analysis of variance was performed with p < 0.05 being significant. Results: Multiple significant differences between PDUs were noted at a pressure of 20 cm H 2 O. The Argyle Sentinel Seal (Sherwood Medical; Tillamore, Ireland) had significantly lower flow rates (mean SD, L/min) compared with all other models. The Argyle Aqua-Seal (Sherwood Medical) had the highest PDU flow rate of devices tested ( L/min). The accuracy of PDUs at manufacturer-suggested settings varied from a mean percentage error of 0.0 to 15.5% from expected pressures; significant differences were noted in accuracy among multiple interdevice pressure comparisons. Similarly, multiple significant flow rate differences between PDCs were noted at 20 cm H 2 O. Lowest flow rates were noted with thoracentesis catheters (used as PDCs) containing side ports. Arrow drainage catheters (14F, pigtail and straight) [Arrow International; Reading, PA] both had significantly greater flow rates (both, L/min), compared with the 14F ( ) and 16F ( ) Cook devices (Cook; Bloomington, IN). Conclusions: These differences in flow rates for PDUs and PDCs may be clinically important, particularly in patients with large pneumothorax-related air leaks. Observed differences in PDU-generated pressures are likely not clinically important. (CHEST 2003; 123: ) Key words: bronchopleural fistula; catheter; chest tube; flow; pneumothorax; pleural drainage unit Abbreviations: ACCP American College of Chest Physicians; ANOVA analysis of variance; BTPS body temperature, ambient pressure, saturated with water vapor; PDC pleural drainage catheter; PDU pleural drainage unit After introduction of the three-bottle, water-seal system for pleural cavity drainage, many commercially available adaptations of this system have *From the Division of Pulmonary and Critical Care Medicine, University of Mississippi Medical Center, Jackson, MS. Presented in part at the American Thoracic Society Annual International Meeting, Toronto, ON, Canada (May 5 10, 2000), and American College of Chest Physicians Annual Meeting, San Francisco, CA (October 21 26, 2000). Manuscript received July 22, 2002; revision accepted December 3, Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( permissions@chestnet.org). Correspondence to: Michael H. Baumann, MD, FCCP, Professor of Medicine, Division of Pulmonary and Critical Care Medicine, University of Mississippi Medical Center, 2500 North State St, Jackson, MS ; mbaumann@medicine. umsmed.edu become available. These commercial pleural drainage units (PDUs) appear to dominate the marketplace to the extent that the original three-bottle system is a rarity in practice today. Portability, ease of use, and disposability are significant advantages of available PDUs; however, the question remains, how reliable are these devices? The only available assessments of commercial PDUs were published in the 1980s, 1,2 noting flow rates for these devices. The four assessed devices are no longer available. No assessment of PDU-generated pressure accuracy has been published. Similarly, no manufacturer-independent analysis of the myriad of commercially available pleural drainage catheters (PDCs) has been published Clinical Investigations

2 Given the recent American College of Chest Physicians (ACCP) recommendations for use of smallbore catheters in certain settings of spontaneous pneumothorax, 3 and their use in traumatic and iatrogenic pneumothoraces, knowledge of the potential air flow rate limitations of these small-bore catheters is key, especially in the setting of a bronchopleural fistula. 4 Limitation in the flow rates accommodated by the PDU or the PDC can lead to the development of a tension pneumothorax. PDU-delivered pressures may also affect device efficiency; hence, knowledge of the flow rate limitations of commercially available PDUs and PDCs and of the accuracy of PDUdelivered pressure would be valuable to the daily management of pneumothoraces. PDU Assessment Materials and Methods Eight PDU models were supplied free of charge by three leading manufacturers (Table 1). These models appear to be the most widely used units in the United States per information supplied by the respective manufacturers. All units are designed to represent the three chambers under water seal concept: a collection chamber, a water-seal chamber, and a suction control chamber. Suction control is achieved either by a mechanical device without need for water (termed dry suction control) or by using a water column. Of the eight units we tested, four units used dry suction control and four units used a water column. RT 200 Calibration Analyzer The RT 200 Calibration Analyzer (Timeter Instrument Corporation; St. Louis, MO) was used for measuring PDU flow rates and pressures. The RT 200 Calibration Analyzer (Fig 1) provides a digital readout of the measurements. A pneumotachometer measures gas flows with a resolution of 0.1 L/min in a range of 5 to 180 L/min with an accuracy of 1% or better of the readout. Pressures from 0 to 250 cm H 2 O are measured with a resolution of 0.1 cm H 2 O with an accuracy of 0.5%, according to the operation/service manual of the manufacturer. Our unit was last calibrated for accuracy by Onsite Calibration Service (Louisville, KY) 1 week prior to our data collection on the PDU. Calibration is suggested by the manufacturer to occur once yearly. This measurement device may be set to make all measurements at body temperature, ambient pressure, saturated with water vapor (BTPS). Measurement Methodology Flow Measurement: A bench set-up was created as in Figure 1. The PDUs were prepared following manufacturer instructions. The suction port of the PDU was connected to wall suction through a regulator. The patient connection tubing of the PDUs was connected to the flow outlet of RT-200 Calibration Analyzer. Extra connection tubing was required for this, and was minimized in length (measured at 15 cm) with a diameter (approximately 2.5 cm) significantly larger than that of the connection tubing supplied with the PDU to avoid additional resistance to flow. The flow inlet of RT-200 Calibration Analyzer was kept open to room air to simulate a continuous large bronchopleural fistula. Each PDU was tested for flow rates at suction control settings of 10 cm H 2 O, 20 cm H 2 O, and 40 cm H 2 O (if available), in that order. Five PDUs were tested for each given model. The PDU models were selected in random order. All five units were tested before proceeding to another model. After initiation of the suction, time was allowed to achieve a steady-state level (the digital readout on the RT-200 Calibrator Analyzer remained constant 0.01) prior to recording the data. All data were recorded at BTPS. Pressure Measurement: The PDUs were set up as above, except the patient connection tubing of each PDU was connected to the pressure measurement port on the RT-200 Calibrator Analyzer. Each PDU was tested for negative pressure generated at suction control settings as for flow measurements. After initiation of the suction, time was allowed to achieve a steadystate level (the digital readout on the RT-200 Calibrator Analyzer remained constant 0.01) prior to recording the data. All data were recorded at BTPS. The accuracy of the device (mean percentage of error) was used for all statistical comparisons of pressures generated by the PDU tested. Accuracy is the standard Table 1 PDU Models, Manufacturer, and Flow Rates* Suction Level, cm H 2 O PDUs Manufacturer/Model Atrium Medical Corporation (Hudson, NH) Atrium Ocean no 40 setting Atrium Atrium 3612 (pediatric device) Sherwood Medical (Tullamore, Ireland) Argyle Aqua-Seal no 40 setting Argyle Thora-Seal III no 40 setting Argyle Sentinel Seal no 40 setting Deknatel (Fall River, MA) Pleur-evac A-6000 (Code #A-6000) Pleur-evac SAHARA (Code #S11000FS) *Data are presented as L/min (mean SD). Water device: pressure regulation by water column. Dry device: pressure regulation not by water column. CHEST / 123 / 6/ JUNE,

3 Figure 1. The set-up to measure PDU flow. Shown (left to right), sequentially connected, are the RT-200 Calibration Analyzer, a PDU, and a wall suction regulator. biomedical instrumentation expression of how well the function of an instrument approaches the true or reference value, 5 in this case the expected delivered negative pressure. A final test of accuracy of pressure was performed after the above measurements were completed. In all other prior pressure accuracy tests outlined above, the directions of the manufacturer regarding the titration of external suction level (wall suction through a regulator) were closely followed. Many wall suction regulators offer the option of full vacuum. This essentially allows the full negative pressure of the suction source to be applied to the PDU. This appears in our experience to be a common practice utilized when setting up a PDU. To test the accuracy of the PDU in down-regulating the full force of the suction source to the pressure set on the PDU, each PDU was retested after all the measurements outlined above were obtained. For each of the pressure settings a PDU was capable of ( 10 cm H 2 O, 20 cm H 2 O, and 40 cm H 2 O, as available), the PDU was retested with the wall suction regulator set to full vacuum and the derived PDU pressure measured as outlined above. Statistical Analysis For both flow and pressure measurements, two-way analysis of variance (ANOVA) with repeated measures was used. Comparison of data at 40 cm H 2 O was made separately from data at 10 cm H 2 O and 20 cm H 2 O, because only four PDUs offered a 40 cm H 2 O suction control level. Post hoc multiple comparison analysis was performed, if appropriate, using a Tukey test. A probability value 0.05 was considered significant. Comparisons were done between all models at 10 cm H 2 O and 20 cm H 2 O, as not all PDUs offered a setting of 40 cm H 2 O. Pressure accuracy was defined as mean percentage error (deviation from the set value SD percentage). PDC Assessment Eleven different catheter types were obtained from three major manufacturers (Arrow International [Reading, PA]; Cook [Bloomington, IN]; and Argyle [Sherwood Medical]) for testing (Table 2). The catheters are listed in the order of ascending expected flow rates based on the bore and length of the catheter as described by the Fanning equation. Briefly, the larger the catheter bore and shorter its length the greater the expected flow accommodated (a complete explanation of the Fanning equation is provided in the Discussion ). Each company donated the needed test samples from everyday stock supplies. Five catheters of each type were tested. Flow Measurement: A bench set-up was created as in Figure 2. As with the PDUs, flow measurements were obtained using the RT 200 Calibration Analyzer. Measurements were made four months after the last calibration. The Pleur-evac A-6000 (Deknatel) PDU with 60 mm Hg of applied wall suction was used for testing of all catheter sets. The Pleur-evac A-6000 was chosen given its combination of suction accuracy and high flow rates, particularly at 20 cm H 2 O, accommodated in a dry unit. A water column unit was not chosen given the potential of variability in the water column height during testing of each PDC group. The testing apparatus consisted of a large carboy connected to the RT 200 Calibration Analyzer, which was then connected in serial to the PDU through which suction was applied. All connections were with rigid tubing with elastic foam sealant tape at all joints to prevent air leaks. The catheters themselves were placed in adjustable adapters supplied by Cook and sealed with silicon caulk. The catheter/adapter assembly was inserted into a large rubber cork, and the cork was inserted into the carboy for testing purposes. The RT 200 Calibration Analyzer was set at BTPS and zeroed prior to testing of each catheter. Three pressure settings ( 10 cm H 2 O, 20 cm H 2 O, and 40 cm H 2 O) were used during testing, and each catheter was occluded prior to each measurement while under suction to ensure the absence of leaks in the apparatus (0 L/min of flow). The catheters in a specific set were all tested at the same negative pressure to decrease the chances for variability. Negative pressures were increased sequentially, and testing of all catheters in each set was completed Clinical Investigations

4 Table 2 PDC Description and Flow Rates* Suction Level, cm H 2 O Model Manufacturer Size Description AK Arrow 8.0F by 16 cm Pneumothorax kit AK Arrow 8.0F by 12 cm Pleura-seal thoracentesis kit (three-way stopcock and selfsealing valve) Argyle Safety Argyle 8.0F by 10 cm Safety pneumothorax system Pneumothorax Argyle Safety Thoracentesis Argyle 8.0F by 8.5 cm Turkel safety thoracentesis system (catheter with sideport/three-way stopcock and self-sealing valve) C-TPTS FSNS Cook 8.5F by 7.5 cm Emergency pneumothorax set C-TPT-100 Cook 9.0F by 29 cm Pneumothorax set C-UPTP-1400-WAYNE Cook 14.0F by 29 cm Pneumothorax set AK Arrow 14.0F by 23 cm Percutaneous cavity drainage catheterization kit (curved catheter with sidearm port) AK Arrow 14.0F by 23 cm Percutaneous cavity drainage catheterization kit (straight catheter with sidearm port) C-TQTS-1600 Cook 16.0F by 41 cm Quick chest tube set C-TQTS-2400 Cook 24.0F by 41 cm Quick chest tube set *Data are presented as L/min (mean SD). Arranged in predicted increasing flow rates as would be predicted by catheter bore and length, those with the smallest bore size and, less importantly, longer length would be expected to have the lower flow rates. After all of the sample sets had been tested, the mean air flow for each product was calculated and used for comparison. Statistical Analysis Two-way ANOVA with repeated measures was used, with models of catheters as the random factor, and the three suction pressures as fixed factors. Post hoc multiple comparison analysis was performed using the Tukey test when differences were found by ANOVA. A probability value 0.05 was considered significant. PDU Assessment Results The PDU mean flow rates achieved at various water pressure suction levels appear in Table 1. Multiple statistically significant differences and areas of similarities are noted. There is nearly a fourfold variation from the PDU with highest flow (Argyle Aqua-Seal) and the PDU with the lowest flow (Argyle Sentinel Seal). The Atrium Ocean 2002, Argyle Thora-Seal III, and Argyle Aqua-Seal provide the highest flow rates, and the units are statistically similar at 10 cm H 2 O and 20 cm H 2 O of suction. The Argyle Sentinel Seal provides the lowest flow rates and was statistically different from all other PDUs at both 10 cm H 2 O and 20 cm H 2 O of pressure. The Pleur-evac A-6000, Pleur-evac SAHARA, Atrium 3600, and Atrium 3612 provide intermediate flow rates. Statistical assessment of those PDUs (n 4) producing 40 cm H 2 O of pressure demonstrates that each model is significantly different from each other (Table 1). The accuracy (mean percent error) of the PDUs tested using the directions of the manufacturers regarding external suction titration appears in Table 3. Multiple interdevice differences in accuracy are significantly different but likely not of clinical significance. Focusing on the PDU 20 cm H 2 O setting frequently incorporated in practice, three major findings are noted when following manufacturer instructions. The Atrium 3612 was the least accurate PDU, demonstrating a mean error of 15.5%, with this accuracy being different from all other PDU models tested at 20 cm H 2 O. The Sentinel Seal was the most accurate PDU, with a mean error of 0.0%, notably with manufacturer instructions necessitating wall suction being set at full vacuum. Accuracy of the other units was variable, but all devices had a mean error of 10%. The errors of all of the PDUs were in the positive direction (delivering less negative pressure than the PDU test value) except for the Thora-Seal III, which demonstrated a mean error of 0.8% in the direction of increased negative pressure from the test (PDU set) value. Regarding the question of the effects of fullvacuum vs manufacturer-directed external suction specifications at the PDU setting of 20 cm H 2 O CHEST / 123 / 6/ JUNE,

5 Figure 2. The set-up to measure PDC flow. Shown (left to right), sequentially connected, are the PDU (Pleur-evac A-6000), the RT-200 Calibration Analyzer, and a carboy with a PDC seated in its top opening. (within the same PDU model), several findings are noted (no data table provided). Most notable, full vacuum applied to any of these PDUs does not appear to create a clinically detrimental increase in negative pressure delivered to the patient. The Pleur-evac A-6000, Atrium Ocean 2002, and the Aqua-Seal produced similar accuracy (comparing same model PDUs to same model PDUs at the two different external suction settings). All other PDU models demonstrated statistically significant pressure accuracy error differences (but not likely clinically significant) between the two external suction settings (manufacturer-suggested wall suction setting PDUs Table 3 Accuracy of Pressures* Suction Level, cm H 2 O Atrium Ocean No 40 setting Atrium Atrium Aqua-Seal NT No 40 setting Thora-Seal III No 40 setting Sentinel Seal No 40 setting Pleur-evac A Pleur-evac SAHARA *Data are presented as mean percent error SD. Accuracy not tested (NT) at 10 cm H 2 O due to technical issues. vs full wall suction) in the negative direction (delivering more negative pressure than the PDU test value), with the Thora-Seal III being least accurate and producing a mean error of 27.0%. The next closest mean error is 8.7% by the Pleur-evac SAHARA. The Sentinel Seal was tested at all PDUset negative suction pressures utilizing external suction source at full vacuum, as necessitated by manufacturer directions, and delivered very accurate pressures (see above). Of the PDUs providing 40 cm H 2 O settings (Table 1), the Pleur-evac A-6000 and Pleur-evac SAHARA produced similar accuracy (comparing same-model PDU to samemodel PDU at the two different external suction settings). The Atrium 3600 and 3612 demonstrated significant pressure accuracy error differences between the two external suction settings in the negative direction (delivering more negative pressure than the PDU test value), with mean accuracy errors of 2.3% and 2.9%, respectively. PDC Assessment Flow rates achieved at various water pressure suction levels for the PDCs tested appear in Table 2. These data may be summarized in five major points: (1) the highest flow rate at all negative pressures is seen with the Cook 24F chest tube; (2) at all pressures, the 14F Arrow drainage catheters had 1882 Clinical Investigations

6 statistically greater flow rates than both the 16F and 14F Cook catheters (approximately 2 L/min and 4 L/min greater, respectively, at 20 cm H 2 O); (3) the majority of catheters (6 of 11 catheters) at all suction levels tested ( 10 cm H 2 O, 20 cm H 2 O, and 40 cm H 2 O) have flow rates 16 L/min; (4) at 20 cm H 2 O (an often-used level of suction pressure in clinical practice), only three catheters (Arrow 14F AK-01600, Arrow 14F AK-01601, and Cook 24F C-TQTS-2400) had flow rates 16 L/min; and (5) the lowest mean flow rates are seen in the two catheters with three-way stopcocks (Argyle Safety Thoracentesis and Arrow AK Pleural-seal Thoracentesis kit). Discussion Chest tube placement and subsequent connection to a PDU may be appropriate for many pneumothoraces. The settings appropriate for chest tube placement, the size of tube selected, including smallerbore tubes, and subsequent utilization of a PDU are well outlined for spontaneous pneumothoraces in the recent ACCP spontaneous pneumothorax consensus statement. 3 Consensus is less defined regarding chest tube utilization, tube size selection, and the role of PDUs in nonspontaneous pneumothoraces, but some evidence provides direction. 6 Key to the selection of a PDC and a PDU in the setting of both spontaneous and nonspontaneous pneumothoraces is the maximal air flow (the volume of air per unit of time) each device accommodates particularly if an air leak (bronchopleural fistula) persists. Gases leaked by a bronchopleural fistula are moist with turbulent flow characteristics. The Fanning equation determines the flow of moist gas with turbulent flow characteristics through a PDC (v 2 r 5 P/fl; v flow, r radius, l length, P pressure, f friction factor) Obviously, the critical factor in chest tube selection is the internal diameter (bore) of the tube and, less so, the tube length. Air flow may be compromised by the presence of viscous fluids such as blood. When considering air drainage alone, tube selection must account not only for removal of existing pleural air, but also the ongoing removal of additional air produced (bronchopleural fistula) and the rate of that air production. Patients with bronchopleural fistulae in the setting of chest trauma, thoracotomy, and ARDS may have air leaks ranging from 1 L/min up to 16 L/min. 2,8,11 Conceivably, patients with spontaneous pneumothoraces, especially those requiring mechanical ventilation, may also have large air leaks. Practical experience indicates that 16 L/min is not a fixed upper limit for air leak volume, with larger leaks possibly occurring in any patient receiving mechanical ventilation with a persistent bronchopleural fistula. The majority of the minute ventilation in an intubated patient receiving mechanical ventilation can exit a large bronchopleural fistula. This may be the case in a postpneumonectomy stump dehiscence. Appropriate PDC size selection is therefore key to preventing tension pneumothorax development. Because a PDU is often connected in series with a PDC, appropriate PDU selection is also key. However, when considering flow rates accommodated, the PDC is the most likely rate-limiting device when compared to the PDU (Tables 1, 2). No formula analogous to the Fanning equation exists to determine flow rates for commercially available PDUs. This is likely due to their relatively complex structure based on the three-bottle (compartment) system. This three-bottle system is commercially condensed into a convenient and easily mobile single module of variable design incorporating the three compartments. The three compartments, sequentially, include the following: (1) the collection bottle to trap liquid material and other debris from the patient s pleural space while allowing pleural air to pass through the next two compartments, (2) the water seal bottle to prevent air flow back to the patient s pleural space and allow detection of an air leak (bronchopleural fistula), and (3) the manometer bottle to regulate the amount of negative pressure transmitted back to the patient from the wall suction device (or equivalent suction source). The manometer bottle has an input and output tube and a central vent tube that may be raised or lowered. Adjusting the depth of the central vent tube in the chamber water determines the negative pressure transmitted from the wall suction back to the patient s pleural space. 9,12 The water chamber device has been replaced in some commercial PDUs with a spring-loaded or similar device to regulate the negative pressure transmitted to the patient. Given the complexity of PDUs and their design variability, it is understandable there is no formula to calculate flow. Meantime, using the Fanning equation to mathematically derive flow rates for a PDC is onerous and impractical. This problem is compounded by the absence of flow rate information in the packaging with either PDUs or PDCs. This absence of information leaves the clinician unable to compare flow characteristics of these devices and objectively choose the optimal device fitting the clinical situation. The absence of packaging information regarding the accuracy of pressures generated by PDUs further complicates the picture. These factors prompted the current study. CHEST / 123 / 6/ JUNE,

7 The multiple flow rate differences among the tested PDCs highlight the need for clinicians to pick the appropriate catheter for specific patient circumstances. This is particularly important in circumstances where high flow rates (persistent large bronchopleural fistulae) may be encountered. The popularity of smaller-bore PDCs, despite some offering potentially dangerous low flow rates, potentially worsens the situation. Contributing to this popularity is the intuitive notion that smaller PDCs, particularly those incorporating the Seldinger technique (guidewire placement), may be less painful to place than the traditional larger-bore tubes often placed by blunt dissection; however, a limited assessment of this issue indicates that smaller-bore catheters placed for simple aspiration of a pneumothorax appear less painful compared to traditional chest tube placement based on their average daily pain score, not from pain during tube placement. 13 The clinician should not hesitate to choose a traditional large-bore chest tube if the situation warrants; however, knowledge of the flow rates of the PDCs tested in this study may allow clinicians to safely select a smaller-bore chest tube. The maximal flow rate accommodated of the PDCs tested at 20 cm H 2 O, as expected, was from the largest-bore PDC tested, the 24F Cook C- TQTS-2400 (Table 2). This catheter had the highest flows at all pressures tested. Catheters 16F may be considered small bore and deliver significantly less flow than the only larger-bore PDC tested, the Cook C-TQTS Various clinical conditions outlined above may have bronchopleural fistula flow rates of up to 16 L/min. 2,8,11 The clinician, however, must be vigilant for air leaks 16 L/min, as noted earlier. The Arrow AK and AK are the only two 16F PDCs tested able to accommodate the arbitrary value of 16 L/min of flow at 20 cm H 2 O. Patients with spontaneous pneumothoraces who are not receiving mechanical ventilation are likely not at high risk of a large bronchopleural fistula and may be appropriate for a 16F PDC. The ACCP spontaneous pneumothorax guidelines suggest initial placement of a chest tube and hospital admission as preferred management of an unstable patient with a large ( 3 cm lung collapse) primary spontaneous pneumothorax, or any secondary spontaneous pneumothorax patient with a large pneumothorax or with clinical instability. Specifically, patients with a primary spontaneous pneumothorax (unlikely at risk for a large air leak) suitable for chest tube placement should have their lung re-expanded using a smallbore catheter ( 14F) or placement of a 16F to 22F chest tube. 3 When to place a chest tube and what size to select is less well defined for nonspontaneous pneumothoraces. A recent review suggests, based on available evidence, placement of larger-bore PDCs (28F to 36F) for noniatrogenic traumatic pneumothoraces, particularly in the patient receiving mechanical ventilation, to drain both an ongoing air leak and potential accompanying hemothorax. 6 Iatrogenic traumatic pneumothoraces may be treated with smaller-bore PDCs as long as patients are not receiving mechanical ventilation. 6 Perhaps in both spontaneous and nonspontaneous pneumothoraces the selection of a PDC with flow rate reserve would be wise to provide a safety net in case of unanticipated increased air flow rates. The premise that larger-bore PDCs should provide larger flow rates breaks down when assessing the Cook 14F (C-UPTP-1400-WAYNE) and 16F (C-TQTS-1600) catheters compared with the Arrow 14F catheters (AK and AK-01601) [Table 2]. Based on the Fanning equation, the clinically significant, lower flow rates of the Cook 14F and 16F catheters at 20 cm H 2 O (approximately 4 L/min and 2 L/min less, respectively) may be accounted for by their greater length (29 cm and 41 cm, respectively) compared to the Arrow 14F catheters (both 23 cm). Conceivably, these differences may also be accounted for by differences in the catheter bore size not reflected by the sizes reported by the manufacturers. Lastly, different PDC construction materials may be more or less collapsible under negative pressure. Ease of collapse would be particularly important in the patient where tissues surrounding the PDC could also drive collapse. PDC collapsibility and its affect on air flow were not assessed. Regardless, differences in air flow rates may not be recognized by a clinician selecting a catheter simply based on bore size. Lastly, using PDC equipment designed as thoracentesis equipment for pneumothorax management may not be a wise choice. Such issues are becoming more prevalent in a cost-containment environment when attempting to purchase one device to meet multiple clinical needs. Both the Argyle and Arrow thoracentesis PDCs have clinically significant less air flow rates compared to their comparable pneumothorax PDCs despite the same bore (each 8F) and the lowest flow rates of all PDCs tested at all tested pressure settings (Table 2). The Argyle thoracentesis PDC accommodates less than half the flow of the comparable Argyle pneumothorax PDC, and the Arrow thoracentesis PDC handles approximately 1.5 L/min less flow than the comparable Arrow pneumothorax PDC (all tests at 20 cm H 2 O). Interestingly, the lengths of both the Argyle and Arrow thoracentesis PDCs are shorter than their 1884 Clinical Investigations

8 comparable pneumothorax catheter of the same internal bore (Table 2); shorter length should enhance not reduce flow rates. The lower flow rates of the thoracentesis catheters may be accounted for by the equipment attached proximally to the thoracentesis catheters including a three-way stopcock and a self-sealing valve. Once a chest tube is placed, whether small or large bore, the tube may then be attached to a PDU. Similar to the selection of a PDC, the appropriate use of a PDU is well delineated in the ACCP spontaneous pneumothorax guidelines and less well defined for traumatic and iatrogenic pneumothoraces. Regardless, if a PDU is connected to a PDC, the same flow considerations exist. Appropriate size selection of a PDC to accommodate bronchopleural fistula air flow may be thwarted by selecting a PDU unable to handle the air flow with subsequent development of a tension pneumothorax. The mean flow rates handled by commercial PDU at 20 cm H 2 O vary considerably, from 10.8 to 42.1 L/min (Table 1). The Sentinel Seal PDU has the lowest flow rate at 10.8 L/min and is lower than any other PDU tested. This average flow is substantially less than may be encountered in various clinical situations, outlined above, and could lead to the development of a tension pneumothorax. Several PDUs deliver 16 L/min (discussed above) at 10 cm H 2 O (Table 1), but all PDUs except the Sentinel Seal deliver 16 L/min at 20 cm H 2 O. The accuracy of the pressures delivered by the PDU varies considerably (Table 3) with multiple interdevice differences; however, the clinical significance of these differences is questionable. For example, the Atrium 3612 was the least accurate PDU, with a mean error of 15.5% at 20 cm H 2 O. This means that the device actually delivers on average 16.9 cm H 2 O( 20 cm H 2 O less 20 cm H 2 O multiplied by 15.5%). This decrease in delivered negative pressure is likely of little clinical significance given the mean flow rate delivered by the device (20.3 L/min, Table 1) at the set PDU pressure of 20 cm H 2 O. Perhaps of more concern would be if the PDU delivered substantially greater negative pressure than indicated. Such elevated pressures theoretically could lead to damage to the underlying tissues including the lung and pericardium. The only device to deliver a negative pressure greater than the indicated PDU set pressure of 20 cm H 2 O (Table 3) is the Thora-Seal III, with a mean increase of only 0.8% (a mean increase in negative pressure of 0.16 cm H 2 O). Accuracy of the other units at the set pressure of 20 cm H 2 O was variable, with a mean error of 10% (other than the Atrium 3612, see above) in the direction of delivering less negative pressure than the PDU test values. Therefore, on average, the delivered negative pressure is reduced by 2cmH 2 O of pressure at the set value of 20 cm H 2 O (a delivered pressure 18 cm H 2 O). These pressure inaccuracies are likely of little clinical significance. Most concern for inappropriate negative pressure delivered to the pleural space might be expected when full-vacuum vs manufacturer-directed external suction specifications are used. Full vacuum does not appear to substantially adversely affect the delivered pressures when PDUs were tested at 20 cm H 2 O settings (no data table). The Pleur-evac A-6000, Atrium Ocean 2002, and the Aqua-Seal produced similar accuracy (comparing same-model PDU to same-model PDU at the two different external suction settings). The Sentinel Seal was tested at all set negative suction levels at full vacuum, based on manufacturer directions, and delivered very accurate pressures. All other PDU models demonstrated significant pressure accuracy error differences between the two external suction settings (manufacturersuggested wall suction setting vs full wall suction) in the negative direction (delivering more negative pressure than the PDU test value). The least accurate PDU at full wall suction was the Thora-Seal III, with a mean error of 27%. This translates to a mean delivered negative water pressure of 25.4 cm H 2 O. Similarly, the most inaccurate PDU with a 40 cm H 2 O setting when tested at full vacuum, the Atrium 3612, produced a mean error of 2.9% in the negative direction (a mean delivered negative pressure of 41.2 cm H 2 O). Such minimal increases in negative pressure are likely of little clinical consequence. No definitive comparison in the accuracy (Table 3) of the water column PDU vs dry PDU systems can be made (Table 1). Absolute mean accuracy (positive or negative error) of the water column units when tested at a set pressure of 20 cm H 2 O varied from a mean error of 0.0% (Sentinel Seal) to 4.6% (Atrium Ocean 2002). The dry unit absolute mean accuracy when tested at 20 cm H 2 O varied from 4.2% (Pleur-evac A-6000) to 15.5% error (Atrium 3612). Several problems exist with this study of PDUs and PDCs. First, although major US manufacturers were contacted, not all commercially available devices were tested. For example, the TRU-CLOSE Thoracic Vent Procedure Tray (Davis and Geck; Wayne, NJ), a PDC, was not tested due to the inability to adapt the device to our testing system. Larger-bore chest tubes, except the Cook C-TQTS- 2400, were not tested given that clinically significant potential flow restrictions were found to exist primarily with small-bore catheters. Products not available in the United States were not tested. Additionwww.chestjournal.org CHEST / 123 / 6/ JUNE,

9 ally, as product lines evolve, the devices reported in this study may no longer be available or be modified and carry the same or different name. Accordingly, the same name may not equate to the same flow and/or accuracy characteristics. This requires the clinician to always query the maximal flow rates that a PDU or PDC delivers and the accuracy of a PDU when incorporating new products in their practice. Lastly, our sham set-up was designed to assess optimal flows from each PDU and PDC, and the PDU accuracy of delivered pressures. This design, however, may not reflect clinical conditions, particularly for a PDC, wherein the pleural space and its non-air contents may affect the flow a particular device may be capable of handling. The presence of the lung or pleural debris such as blood may significantly decrease the air flows reported from our test design. In conclusion, multiple different flow rates are noted among different PDUs at the same negative water pressure settings. Although significant accuracy differences in delivered pressures of the PDU tested are seen, these may not be of clinical consequence. PDUs, in general, significantly protect the patient (pleural space) in the event of inadvertent use of full wall suction. Similarly, multiple significant differences in PDC flow rates exist, including between catheters of the same bore size. The significant differences in accommodated air flows may be particularly important in patients with large air leaks. Incorporation of such knowledge may preclude the development of a tension pneumothorax. References 1 Capps JS, Tyler ML, Rusch VW, et al. Potential of chest drainage units to evacuate broncho-pleural air leaks. Chest 1985; 88:57S 2 Rusch VW, Capps JS, Tyler ML, et al. The performance of four pleural drainage systems in an animal model of bronchopleural fistula. Chest 1988; 93: Baumann MH, Strange C, Heffner JE, et al. Management of spontaneous pneumothorax: an American College of Chest Physicians Delphi consensus statement. Chest 2001; 119: Baumann MH, Sahn SA. Medical management and therapy of bronchopleural fistulas in the mechanically ventilated patient. Chest 1990; 97: Latman NS, Lanier R. Expressions of accuracy in the evaluation of biomedical instrumentation. Biomed Instrum Technol 1998; 32: Baumann MH. Non-spontaneous pneumothorax. In: Light RW, Lee YCG, eds. Pleural disease: an international textbook. London, UK: Arnold Publishers, 2003 (in press) 7 Baumann MH, Strange C. Treatment of spontaneous pneumothorax: a more aggressive approach? Chest 1997; 112: Batchelder TL, Morris KA. Critical factors in determining adequate pleural drainage in both the operated and nonoperated chest. Am Surg 1962; 28: Miller KS, Sahn SA. Chest tubes: indications, technique, management and complications. Chest 1987; 91: Swenson EW, Birath G, Ahbeck A. Resistance to air flow in bronchospirometric catheters. J Thorac Surg 1957; 33: Powner DJ, Cline D, Rodman GH. Effect of chest-tube suction on gas flow through a bronchopleural fistula. Crit Care Med 1985; 13: Baumann MH. Chest tubes. In: Bouros D, ed. Pleural disease. New York, NY: Marcel Dekker, 2003 (in press) 13 Harvey J, Prescott RJ. Simple aspiration versus intercostal tube drainage for spontaneous pneumothorax in patients with normal lungs. BMJ 1994; 309: Clinical Investigations