- Email: [email protected]
Developing an experimental model for surgical drainage investigations: an initial report
The American Journal of Surgery (2012) 203, 388 –391
Midwest Surgical Association
Developing an experimental model for surgical drainage investigations: an initial report Andrew L. Swartz, B.Sc., Ogochukwu Azuh, M.D., Leila V. Obeid, B.Sc., Anthony J. Munaco, M.D., Shahab Toursavadkohi, M.D., James Adams, B.Ed., Mark Dulchavsky, Liz Dobie, M.P.H., Daniel J. Berardo, Matilda Horst, M.D., J.H. Patton, M.D., Anthony J. Falvo, D.O., Ilan Rubinfeld, M.D.* Department of Surgery, Henry Ford Hospital, 2799 W. Grand Blvd., Detroit, MI 48202, USA KEYWORDS: Jackson-Pratt drains; Closed suction drains; Surgical drains; Abdominal drainage
Abstract BACKGROUND: We sought to pilot and initiate validation of a surgical drainage model. METHODS: We designed a laboratory model to compare Jackson-Pratt surgical drains using 3 soups to emulate body fluids of serous, purulent, and necrotic debris. Each drain was trialed with each of the 3 fluids. Time and completeness of drainage were recorded. A survey of surgical residents and faculty was performed for convenience sampling. RESULTS: Under serous conditions, the round Jackson-Pratt drained the cavity quicker, but left a larger residual volume of fluid. Under purulent conditions, the round Jackson-Pratt was slower and drained less fluid. With debris fluid, the round Jackson-Pratt was quicker with less residual fluid whereas the flat type clogged each time. Survey results showed adequate concordance with surgeons in agreement on soup choice. CONCLUSIONS: The Jackson-Pratt drains perform differently depending on the drainage situation. The surgical community requires improved drain data to drive practice patterns. © 2012 Elsevier Inc. All rights reserved.
Surgical drains have a long history in medical practice dating back to Hippocrates (460 –377 BC) using drains to treat empyema.1,2 No consensus currently exists among surgeons regarding the use of surgical drains. These are high-preference items in surgical practice and lead to much controversy at local and national conferences, as well as in No member of our research team has any financial obligations to any company product used in our study. No member of our research team derives personal profit or gain, directly or indirectly, by sales or promotion of the products used in our study. * Corresponding author: Tel.: ⫹1-313-916-9765; fax: ⫹1-313-9168007. E-mail address: [email protected] Manuscript received July 24, 2011; revised manuscript September 18, 2011
0002-9610/$ - see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2011.09.015
the context of morbidity and mortality discussions. Controversy persists related to type, kind, size, and duration of use, with little data to drive these passionate discussions.3– 8 Our institutional and departmental morbidity and mortality, as well as sentinel event committees, have identified the round and flat Jackson-Pratt drains (Cardinal Health, Dublin, OH) as a potential source of unnecessary variability. The Jackson-Pratt drains have had documented incidences of adverse events. A search through the US Food and Drug Administration’s Manufacturer and User Facility Device Experience database (1992–2011) revealed several reports about both types of Jackson-Pratt drains breaking during attempted removal from patients, thus requiring surgical re-exploration for retrieval. While analyzing the US Food and Drug Administration’s Manufacturer and User
A.L. Swartz et al.
Surgical drainage investigations
Facility Device Experience data, we noticed a higher rate of sentinel events occurring with the flat Jackson-Pratt drains, based on the transition point. Because these drains are the most common drains used in abdominal surgery,9 –11 this safety data led us to evaluate the 2 types of Jackson-Pratt surgical drains using a laboratory model for objective testing with reproducible laboratory analogues representing the major body fluid types.
Methods Our first need was a cavitary model. We had no generic laboratory model previously established for drainage, despite extensive literature review, and we wished to avoid compromising human beings or animals because our initial focus was on the dynamics of the drain and the fluids. Thus, we began developing a model for drainage of a conceptual cavity such as the lesser sac. Ballistic gelatin, which has been used by law enforcement and forensic experts to simulate density and viscosity of living human tissue,12 is convenient, reproducible, and can be re-created easily at other institutions. It is a simple bowl-shaped cavity with a ballistic gel lid. The shape of the cavity can be open and resealed for ease of trials and reproducibility; it is not collapsible. We created 2 standardized ballistic gel cavities using ballistics powder, 2 tin containers, and an electric oven. Once the cavities fully formed and cooled, a scalpel was used to puncture one of the sides to insert both the fluid and surgical drain. To create a closed suction system, the Jackson-Pratt 200-mL (JP200) suction bulb (Cardinal Health, Dublin, OH) was connected to the free end of the drain. Once attached, the bulb was depressed fully and sealed. When the bulb was ready, it was dropped and the handheld timer was started. Once the end point was reached (a fully expanded bulb) the timer was stopped, and the final drainage time was recorded. The bulb drain was emptied into a measuring cup, and the volume of soup evacuated and the residual volume of fluid left in the ballistics gel cavity were recorded. Once the values were noted, the drain and residual soup were removed by creating a small slit in the side of the cavity. The gel then was resealed using a propane torch. Once the cavity and JP100 bulb were emptied, another trial was run. These were short-duration drainage episodes, not overnight or multiday drainage scenarios. Similarly, we had no literature-based nonanimal or human model for drainable fluids. We hypothesized that certain canned soups could mimic the liquid composition of various bodily fluids. Campbell’s chicken broth, pea, and Italian Wedding soups (Campbell Soup Company, Camden, NJ) were used to emulate serous fluid, purulent fluid, and fluid with debris, respectively. A large syringe, without a tip, was used to collect 150 mL of one type of soup and injected into the cavity. Once the cavities were loaded with 150 mL of soup, 1 of the 2 surgical drains (flat or round
389 Jackson-Pratt) was inserted into the cavity. The entry site into the cavity was then sealed airtight with extra ballistic gel and a handheld propane torch. To justify having different types of drains, we needed to show at least a 10% difference in drainage characteristics. Based on a power analysis, 20 trials per drain was sufficient. In total, 120 trials (20 trials per drain, per each soup) were conducted using bulb suction. The results (time [sec], volume removed by bulb suction [mL], and residual volume left in cavity [mL]) were analyzed using Minitab 16 (Minitab, Inc. State College, PA). In conjunction with the laboratory model, approval from the institutional review board was obtained. Because there was no pre-existing reproducible drainage models based on scientific or engineering principles (ie, viscosity index), validation of our choice was sought by convenience sampling of the surgical community through a survey. Convenience sampling was appropriate in this instance because we sought to develop a multi-institutional base of survey results crossing education, generational, and institutional biases. The survey was administered only to surgeons not involved with the research at regional surgical meetings and represents a broad view for face and content validity. The survey gathered demographic data, and each soup type was assessed as a potential analogue for a drainable fluid. Results were tabulated and analyzed in SPSS (IBM, Armonk, NY).
Results Under isolated conditions (serous, purulent, and fluid with debris) independent sample t test analysis revealed that the round Jackson-Pratt drain was superior to the flat type with regard to time (mean difference, ⫺1.32 s; P ⬍ .001), but not in terms of residual volume (mean difference, 2.2 mL; P ⫽ .124) for serous conditions (chicken broth). Under purulent conditions (pea soup), analyses showed that the round drain was inferior to the flat drain in terms of time (mean difference, 186 s; P ⬍ .001) and residual volume (mean difference, 4.9 mL; P ⬍ .001). Analysis using the fluid that contained debris (Italian Wedding soup) showed that the round Jackson-Pratt was superior in terms of both time (mean difference, ⫺25.2 s; P ⫽ .004) and residual volume (mean difference, ⫺109.2 mL; P ⬍ .001). Independent sample t test analysis of the pooled data across all categories resulted in a significant difference between the mean time (mean difference, 52.8 s; P ⫽ .027) and residual volume (mean difference, ⫺34.0 mL; P ⬍ .001) of the round and flat Jackson-Pratt drains. Table 1 shows the average time, average residual volume, and average volume drained for each drain fluid type. Survey results totaled 84 respondents: 43% faculty, 57% residents. The residents were spread evenly across postgraduate years (PGY) 1 to 5 (PGY 1, 11.9%; PGY 2, 7.1%; PGY 3, 8.3%; PGY 4, 16.7%; and PGY 5, 13.1%). Of all respondents, 73% self-reported being from academic centers and
390 Table 1
The American Journal of Surgery, Vol 203, No 3, March 2012 Drain averages
Soup type
Jackson-Pratt type
Average time, s
Chicken broth
Round Flat Round Flat Round Flat
18 20 298 300 326 138
Italian wedding Pea
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
1.6 1.2 35.9 .0 22.0 11.4
overall 82% always had residents involved in the cases. Most respondents were general surgery residents and faculty (83.3%). Drain preference questions, which allowed respondents to choose between fluid and drain type, showed no significant differences between residents and faculty. Of the 4 questions related to clinical scenarios and drain choice, no statistically significant differences were noted in drain use between residents and faculty (P values were as follows: Serous, .44; Purulent, .17; Bloody, .26; Necrotic debris, .75). However, 2 questions on drain attributes with evidence-based correct answers, had statistically significant differences because no faculty selected the incorrect choice. Residents preferred open drains (17.8% vs 0% faculty; P ⫽ .008) and sump drainage (15.2% vs 0% faculty; P ⫽ .038). Inter-rater reliability for the scenario usage questions also was better between faculty, with the Cronbach ␣ at .724 versus .629 for the residents. Inter-rater reliability was perfect among the faculty on the 2 content questions of note.
Conclusions To date, little quantitative data exists comparing the performance and/or complications of surgical drains, and current literature appears confounding and/or subjective. We aimed to develop a straightforward, reproducible laboratory model that would provide statistical evidence to support or refute the claims made not only by the current literature, but also by surgeons at our and other institutions. The performance of each of the surgical drains (round and flat Jackson-Pratt) varied in the different laboratory scenarios. Under serous conditions (chicken broth), both the round and flat Jackson-Pratt drains emptied the cavity in mere seconds, averaging 18 and 20 seconds, respectively. The round type drained the cavity quicker, but left a larger volume of fluid behind. An explanation could be that the lumen size of the perforated end of the round Jackson-Pratt drain is larger than that of the flat drain, therefore making it able to accommodate a greater amount of fluid. The round Jackson-Pratt drain might have left more fluid in the gel model because it covers less surface area of the bottom of the gel cavity as opposed to the flat drain with its wide perforated end. For purulent conditions (pea soup), both drains evacuated the cavity at a much slower rate than under the serous conditions, averaging 326 seconds for the round drain and
Average volume drain, mL 103.85 106.05 109.4 .2 101.3 106.2
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.66 5.61 10.4 .52 2.11 4.11
Average residual volume, mL 46.15 43.95 40.6 149.8 48.7 43.8
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.66 5.61 10.4 .52 2.11 4.11
138 seconds for the flat drain. This finding was expected because of the increased viscosity of the fluid. However, the round Jackson-Pratt drain was both slower and drained less fluid. Further investigation is needed to determine why the round drain emptied the cavity at a slower rate, but we suspect the reason may be that the round Jackson-Pratt drain covers less surface area of the bottom of the cavity when compared with the flat drain. Under the conditions in which the fluid contained debris (Italian Wedding soup), the round Jackson-Pratt drain was superior in terms of both time and residual volume. The round drain was able to accommodate the small bit of debris in the soup, and drained with an average time of 298 seconds. The flat drain, however, clogged during each of the 20 runs, and time was always stopped at an end point of 5 minutes. We were only able to collect fluid in the JP100 bulb during the first 3 runs with the flat Jackson-Pratt drain (2 mL, 1 mL, and 1 mL, respectively); the remainder of the trials failed to drain any liquid. We attribute this failure of the flat drain to its narrow lumen at the perforated end, which always became obstructed with small debris. This study had multiple limitations and sources of bias. In two thirds of the trials performed by Whitson et al,3 the JP100 bulb failed to collect fluid well before reaching its 100-mL volume, this happened in our experience as well. Our model was not a true replication of a fibrin-rich neovascular abscess. The ballistic gelatin cavity lacked the biophysical structure of living tissue that would allow it to conform and/or collapse during drainage. We drained for a relatively short time period. The fluids were not actually biologically real fluids. The cavity may not adequately reflect a real drainable environment. Finally, human error may have occurred within our trials, such as a slight delay in pressing the start/stop button on our timer. The survey results were adequately unbiased based on diversity of training and institutions. More importantly, there was reasonable concordance regarding soup and fluid comparison based on the survey. Overall we found that, under differing conditions, the performance of the 2 surgical drains differed. These results suggest that the difference between the round and flat JacksonPratt drains is not as great as we originally presumed. Future studies aiming to evaluate these 2 common surgical drains might use experimental differences such as different types of bulbs, for example, the Jackson-Pratt 400 mL or the Surgidyne
A.L. Swartz et al.
Surgical drainage investigations
100 mL (Surgidyne, Caledonia, MI); different kinds of fluids, that is, including a component of motion to replicate body fluid dynamic movement; and different physical characteristics of the cavity wall, allowing it to conform/collapse during drainage.
References 1. Gaines RJ, Dunbar RP. The use of surgical drains in orthopedics. Orthopedics 2008;31:702–5. 2. Robb H. The management of the drainage tube in abdominal surgery. Johns Hopkins Hosp Rep 1891;2:184. 3. Whitson BA, Richardson E, Iaizzo PA, et al. Not every bulb is a rose: a functional comparison of bulb suction devices. J Surg Res 2009;156: 270 –3. 4. Barton A, Blitz M, Callahan D, et al. Early removal of postmastectomy drains is not beneficial: results from a halted randomized controlled trial. Am J Surg 2006;191:652– 6. 5. Aitken DR, Hunsaker R, James AG. Prevention of seromas following mastectomy and axillary dissection. Surg Gynecol Obstet 1984;158: 327–30. 6. Santarius T, Kirkpatrick PJ, Ganesan D, et al. Use of drains versus no drains after burr-hole evacuation of chronic subdural haematoma: a randomised controlled trial. Lancet 2009;374:1067–73. 7. Cerise EJ, Pierce WA, Diamond DL. Abdominal drains: their role as a source of infection following splenectomy. Ann Surg 1970;171: 764 –9. 8. Maingot R. Abdominal drainage tubes. Postgrad Med J 1932;8:126 –9. 9. Sakakura N, Fukui T, Mori S, et al. Fluid drainage and air evacuation characteristics of Blake and conventional drains used after pulmonary resection. Ann Thorac Surg 2009;87:1539 – 45. 10. Abramson DJ. Improved triple lumen all purpose drains and their care and management. Am Surg 1983;49:539 – 41. 11. Cardinal Health; 2007. Available at: http://cardinalhealth.com/us/en/ brands/medivac/files/JackPratt%20Brochure%20final%202007.pdf. Accessed: July 13, 2011. 12. Nicholas NC, Welsch JR. Institute for Non-Lethal Defense Technologies: Ballistic Gelatin. Institute for Non-Lethal Defense Technologies Report, The Pennsylvania State University, Applied Research Laboratory, The Pennsylvania State University. 2004. Available at: http:// www.nldt.org/documents/ballistic_gelatin_report.pdf.
391
Discussion Dr Donald W. Moorman (Pittsburgh, PA): Replicating biology is very difficult. Is the cavity truly a replication of a fibrin-rich neovascular abscess wall in both elasticity and ability to conform to the internal abscess cavity as it collapses? I did not find it too surprising that chunks in Italian wedding soup plugged the drains, but this really speaks to the validity of your model. Can you clarify the decision to use soups rather than create particulate-specific and viscosity-specific substrates? Was the survey a previously validated survey? Lastly, how did you avoid the biases and limitations of convenience sampling at regional meetings? Dr Ogochukwu Azuh (Detroit, MI): It was not a perfect model and modeling of the intra-abdominal cavity and whatnot, was not necessarily our aim. We analyzed the physical characteristics of the fluid versus the drain. We could repeat this experiment maybe putting the drain in a plastic bag where it can allow compressibility of the size and actually mirror human body drain dynamics. To answer your second question, soup is cheap and easily reproducible and allows for the organic variations of certain tissue types, like chicken broth for serous fluid, and green pea soup for purulent fluid. We elicited the help of a survey expert and passed it out at conferences such as this with the aim of validating our soup model. We did validate this with an inter-rater reliability using Cronbach ␣ analysis. To avoid bias, we performed the survey without background insight as to what we planned to do. We just said, “Hey, listen, could you just answer these questions? What did you think of these different types of soup in order to emulate these different kinds of bodily fluids?”
Midwest Surgical Association
Developing an experimental model for surgical drainage investigations: an initial report Andrew L. Swartz, B.Sc., Ogochukwu Azuh, M.D., Leila V. Obeid, B.Sc., Anthony J. Munaco, M.D., Shahab Toursavadkohi, M.D., James Adams, B.Ed., Mark Dulchavsky, Liz Dobie, M.P.H., Daniel J. Berardo, Matilda Horst, M.D., J.H. Patton, M.D., Anthony J. Falvo, D.O., Ilan Rubinfeld, M.D.* Department of Surgery, Henry Ford Hospital, 2799 W. Grand Blvd., Detroit, MI 48202, USA KEYWORDS: Jackson-Pratt drains; Closed suction drains; Surgical drains; Abdominal drainage
Abstract BACKGROUND: We sought to pilot and initiate validation of a surgical drainage model. METHODS: We designed a laboratory model to compare Jackson-Pratt surgical drains using 3 soups to emulate body fluids of serous, purulent, and necrotic debris. Each drain was trialed with each of the 3 fluids. Time and completeness of drainage were recorded. A survey of surgical residents and faculty was performed for convenience sampling. RESULTS: Under serous conditions, the round Jackson-Pratt drained the cavity quicker, but left a larger residual volume of fluid. Under purulent conditions, the round Jackson-Pratt was slower and drained less fluid. With debris fluid, the round Jackson-Pratt was quicker with less residual fluid whereas the flat type clogged each time. Survey results showed adequate concordance with surgeons in agreement on soup choice. CONCLUSIONS: The Jackson-Pratt drains perform differently depending on the drainage situation. The surgical community requires improved drain data to drive practice patterns. © 2012 Elsevier Inc. All rights reserved.
Surgical drains have a long history in medical practice dating back to Hippocrates (460 –377 BC) using drains to treat empyema.1,2 No consensus currently exists among surgeons regarding the use of surgical drains. These are high-preference items in surgical practice and lead to much controversy at local and national conferences, as well as in No member of our research team has any financial obligations to any company product used in our study. No member of our research team derives personal profit or gain, directly or indirectly, by sales or promotion of the products used in our study. * Corresponding author: Tel.: ⫹1-313-916-9765; fax: ⫹1-313-9168007. E-mail address: [email protected] Manuscript received July 24, 2011; revised manuscript September 18, 2011
0002-9610/$ - see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2011.09.015
the context of morbidity and mortality discussions. Controversy persists related to type, kind, size, and duration of use, with little data to drive these passionate discussions.3– 8 Our institutional and departmental morbidity and mortality, as well as sentinel event committees, have identified the round and flat Jackson-Pratt drains (Cardinal Health, Dublin, OH) as a potential source of unnecessary variability. The Jackson-Pratt drains have had documented incidences of adverse events. A search through the US Food and Drug Administration’s Manufacturer and User Facility Device Experience database (1992–2011) revealed several reports about both types of Jackson-Pratt drains breaking during attempted removal from patients, thus requiring surgical re-exploration for retrieval. While analyzing the US Food and Drug Administration’s Manufacturer and User
A.L. Swartz et al.
Surgical drainage investigations
Facility Device Experience data, we noticed a higher rate of sentinel events occurring with the flat Jackson-Pratt drains, based on the transition point. Because these drains are the most common drains used in abdominal surgery,9 –11 this safety data led us to evaluate the 2 types of Jackson-Pratt surgical drains using a laboratory model for objective testing with reproducible laboratory analogues representing the major body fluid types.
Methods Our first need was a cavitary model. We had no generic laboratory model previously established for drainage, despite extensive literature review, and we wished to avoid compromising human beings or animals because our initial focus was on the dynamics of the drain and the fluids. Thus, we began developing a model for drainage of a conceptual cavity such as the lesser sac. Ballistic gelatin, which has been used by law enforcement and forensic experts to simulate density and viscosity of living human tissue,12 is convenient, reproducible, and can be re-created easily at other institutions. It is a simple bowl-shaped cavity with a ballistic gel lid. The shape of the cavity can be open and resealed for ease of trials and reproducibility; it is not collapsible. We created 2 standardized ballistic gel cavities using ballistics powder, 2 tin containers, and an electric oven. Once the cavities fully formed and cooled, a scalpel was used to puncture one of the sides to insert both the fluid and surgical drain. To create a closed suction system, the Jackson-Pratt 200-mL (JP200) suction bulb (Cardinal Health, Dublin, OH) was connected to the free end of the drain. Once attached, the bulb was depressed fully and sealed. When the bulb was ready, it was dropped and the handheld timer was started. Once the end point was reached (a fully expanded bulb) the timer was stopped, and the final drainage time was recorded. The bulb drain was emptied into a measuring cup, and the volume of soup evacuated and the residual volume of fluid left in the ballistics gel cavity were recorded. Once the values were noted, the drain and residual soup were removed by creating a small slit in the side of the cavity. The gel then was resealed using a propane torch. Once the cavity and JP100 bulb were emptied, another trial was run. These were short-duration drainage episodes, not overnight or multiday drainage scenarios. Similarly, we had no literature-based nonanimal or human model for drainable fluids. We hypothesized that certain canned soups could mimic the liquid composition of various bodily fluids. Campbell’s chicken broth, pea, and Italian Wedding soups (Campbell Soup Company, Camden, NJ) were used to emulate serous fluid, purulent fluid, and fluid with debris, respectively. A large syringe, without a tip, was used to collect 150 mL of one type of soup and injected into the cavity. Once the cavities were loaded with 150 mL of soup, 1 of the 2 surgical drains (flat or round
389 Jackson-Pratt) was inserted into the cavity. The entry site into the cavity was then sealed airtight with extra ballistic gel and a handheld propane torch. To justify having different types of drains, we needed to show at least a 10% difference in drainage characteristics. Based on a power analysis, 20 trials per drain was sufficient. In total, 120 trials (20 trials per drain, per each soup) were conducted using bulb suction. The results (time [sec], volume removed by bulb suction [mL], and residual volume left in cavity [mL]) were analyzed using Minitab 16 (Minitab, Inc. State College, PA). In conjunction with the laboratory model, approval from the institutional review board was obtained. Because there was no pre-existing reproducible drainage models based on scientific or engineering principles (ie, viscosity index), validation of our choice was sought by convenience sampling of the surgical community through a survey. Convenience sampling was appropriate in this instance because we sought to develop a multi-institutional base of survey results crossing education, generational, and institutional biases. The survey was administered only to surgeons not involved with the research at regional surgical meetings and represents a broad view for face and content validity. The survey gathered demographic data, and each soup type was assessed as a potential analogue for a drainable fluid. Results were tabulated and analyzed in SPSS (IBM, Armonk, NY).
Results Under isolated conditions (serous, purulent, and fluid with debris) independent sample t test analysis revealed that the round Jackson-Pratt drain was superior to the flat type with regard to time (mean difference, ⫺1.32 s; P ⬍ .001), but not in terms of residual volume (mean difference, 2.2 mL; P ⫽ .124) for serous conditions (chicken broth). Under purulent conditions (pea soup), analyses showed that the round drain was inferior to the flat drain in terms of time (mean difference, 186 s; P ⬍ .001) and residual volume (mean difference, 4.9 mL; P ⬍ .001). Analysis using the fluid that contained debris (Italian Wedding soup) showed that the round Jackson-Pratt was superior in terms of both time (mean difference, ⫺25.2 s; P ⫽ .004) and residual volume (mean difference, ⫺109.2 mL; P ⬍ .001). Independent sample t test analysis of the pooled data across all categories resulted in a significant difference between the mean time (mean difference, 52.8 s; P ⫽ .027) and residual volume (mean difference, ⫺34.0 mL; P ⬍ .001) of the round and flat Jackson-Pratt drains. Table 1 shows the average time, average residual volume, and average volume drained for each drain fluid type. Survey results totaled 84 respondents: 43% faculty, 57% residents. The residents were spread evenly across postgraduate years (PGY) 1 to 5 (PGY 1, 11.9%; PGY 2, 7.1%; PGY 3, 8.3%; PGY 4, 16.7%; and PGY 5, 13.1%). Of all respondents, 73% self-reported being from academic centers and
390 Table 1
The American Journal of Surgery, Vol 203, No 3, March 2012 Drain averages
Soup type
Jackson-Pratt type
Average time, s
Chicken broth
Round Flat Round Flat Round Flat
18 20 298 300 326 138
Italian wedding Pea
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
1.6 1.2 35.9 .0 22.0 11.4
overall 82% always had residents involved in the cases. Most respondents were general surgery residents and faculty (83.3%). Drain preference questions, which allowed respondents to choose between fluid and drain type, showed no significant differences between residents and faculty. Of the 4 questions related to clinical scenarios and drain choice, no statistically significant differences were noted in drain use between residents and faculty (P values were as follows: Serous, .44; Purulent, .17; Bloody, .26; Necrotic debris, .75). However, 2 questions on drain attributes with evidence-based correct answers, had statistically significant differences because no faculty selected the incorrect choice. Residents preferred open drains (17.8% vs 0% faculty; P ⫽ .008) and sump drainage (15.2% vs 0% faculty; P ⫽ .038). Inter-rater reliability for the scenario usage questions also was better between faculty, with the Cronbach ␣ at .724 versus .629 for the residents. Inter-rater reliability was perfect among the faculty on the 2 content questions of note.
Conclusions To date, little quantitative data exists comparing the performance and/or complications of surgical drains, and current literature appears confounding and/or subjective. We aimed to develop a straightforward, reproducible laboratory model that would provide statistical evidence to support or refute the claims made not only by the current literature, but also by surgeons at our and other institutions. The performance of each of the surgical drains (round and flat Jackson-Pratt) varied in the different laboratory scenarios. Under serous conditions (chicken broth), both the round and flat Jackson-Pratt drains emptied the cavity in mere seconds, averaging 18 and 20 seconds, respectively. The round type drained the cavity quicker, but left a larger volume of fluid behind. An explanation could be that the lumen size of the perforated end of the round Jackson-Pratt drain is larger than that of the flat drain, therefore making it able to accommodate a greater amount of fluid. The round Jackson-Pratt drain might have left more fluid in the gel model because it covers less surface area of the bottom of the gel cavity as opposed to the flat drain with its wide perforated end. For purulent conditions (pea soup), both drains evacuated the cavity at a much slower rate than under the serous conditions, averaging 326 seconds for the round drain and
Average volume drain, mL 103.85 106.05 109.4 .2 101.3 106.2
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.66 5.61 10.4 .52 2.11 4.11
Average residual volume, mL 46.15 43.95 40.6 149.8 48.7 43.8
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.66 5.61 10.4 .52 2.11 4.11
138 seconds for the flat drain. This finding was expected because of the increased viscosity of the fluid. However, the round Jackson-Pratt drain was both slower and drained less fluid. Further investigation is needed to determine why the round drain emptied the cavity at a slower rate, but we suspect the reason may be that the round Jackson-Pratt drain covers less surface area of the bottom of the cavity when compared with the flat drain. Under the conditions in which the fluid contained debris (Italian Wedding soup), the round Jackson-Pratt drain was superior in terms of both time and residual volume. The round drain was able to accommodate the small bit of debris in the soup, and drained with an average time of 298 seconds. The flat drain, however, clogged during each of the 20 runs, and time was always stopped at an end point of 5 minutes. We were only able to collect fluid in the JP100 bulb during the first 3 runs with the flat Jackson-Pratt drain (2 mL, 1 mL, and 1 mL, respectively); the remainder of the trials failed to drain any liquid. We attribute this failure of the flat drain to its narrow lumen at the perforated end, which always became obstructed with small debris. This study had multiple limitations and sources of bias. In two thirds of the trials performed by Whitson et al,3 the JP100 bulb failed to collect fluid well before reaching its 100-mL volume, this happened in our experience as well. Our model was not a true replication of a fibrin-rich neovascular abscess. The ballistic gelatin cavity lacked the biophysical structure of living tissue that would allow it to conform and/or collapse during drainage. We drained for a relatively short time period. The fluids were not actually biologically real fluids. The cavity may not adequately reflect a real drainable environment. Finally, human error may have occurred within our trials, such as a slight delay in pressing the start/stop button on our timer. The survey results were adequately unbiased based on diversity of training and institutions. More importantly, there was reasonable concordance regarding soup and fluid comparison based on the survey. Overall we found that, under differing conditions, the performance of the 2 surgical drains differed. These results suggest that the difference between the round and flat JacksonPratt drains is not as great as we originally presumed. Future studies aiming to evaluate these 2 common surgical drains might use experimental differences such as different types of bulbs, for example, the Jackson-Pratt 400 mL or the Surgidyne
A.L. Swartz et al.
Surgical drainage investigations
100 mL (Surgidyne, Caledonia, MI); different kinds of fluids, that is, including a component of motion to replicate body fluid dynamic movement; and different physical characteristics of the cavity wall, allowing it to conform/collapse during drainage.
References 1. Gaines RJ, Dunbar RP. The use of surgical drains in orthopedics. Orthopedics 2008;31:702–5. 2. Robb H. The management of the drainage tube in abdominal surgery. Johns Hopkins Hosp Rep 1891;2:184. 3. Whitson BA, Richardson E, Iaizzo PA, et al. Not every bulb is a rose: a functional comparison of bulb suction devices. J Surg Res 2009;156: 270 –3. 4. Barton A, Blitz M, Callahan D, et al. Early removal of postmastectomy drains is not beneficial: results from a halted randomized controlled trial. Am J Surg 2006;191:652– 6. 5. Aitken DR, Hunsaker R, James AG. Prevention of seromas following mastectomy and axillary dissection. Surg Gynecol Obstet 1984;158: 327–30. 6. Santarius T, Kirkpatrick PJ, Ganesan D, et al. Use of drains versus no drains after burr-hole evacuation of chronic subdural haematoma: a randomised controlled trial. Lancet 2009;374:1067–73. 7. Cerise EJ, Pierce WA, Diamond DL. Abdominal drains: their role as a source of infection following splenectomy. Ann Surg 1970;171: 764 –9. 8. Maingot R. Abdominal drainage tubes. Postgrad Med J 1932;8:126 –9. 9. Sakakura N, Fukui T, Mori S, et al. Fluid drainage and air evacuation characteristics of Blake and conventional drains used after pulmonary resection. Ann Thorac Surg 2009;87:1539 – 45. 10. Abramson DJ. Improved triple lumen all purpose drains and their care and management. Am Surg 1983;49:539 – 41. 11. Cardinal Health; 2007. Available at: http://cardinalhealth.com/us/en/ brands/medivac/files/JackPratt%20Brochure%20final%202007.pdf. Accessed: July 13, 2011. 12. Nicholas NC, Welsch JR. Institute for Non-Lethal Defense Technologies: Ballistic Gelatin. Institute for Non-Lethal Defense Technologies Report, The Pennsylvania State University, Applied Research Laboratory, The Pennsylvania State University. 2004. Available at: http:// www.nldt.org/documents/ballistic_gelatin_report.pdf.
391
Discussion Dr Donald W. Moorman (Pittsburgh, PA): Replicating biology is very difficult. Is the cavity truly a replication of a fibrin-rich neovascular abscess wall in both elasticity and ability to conform to the internal abscess cavity as it collapses? I did not find it too surprising that chunks in Italian wedding soup plugged the drains, but this really speaks to the validity of your model. Can you clarify the decision to use soups rather than create particulate-specific and viscosity-specific substrates? Was the survey a previously validated survey? Lastly, how did you avoid the biases and limitations of convenience sampling at regional meetings? Dr Ogochukwu Azuh (Detroit, MI): It was not a perfect model and modeling of the intra-abdominal cavity and whatnot, was not necessarily our aim. We analyzed the physical characteristics of the fluid versus the drain. We could repeat this experiment maybe putting the drain in a plastic bag where it can allow compressibility of the size and actually mirror human body drain dynamics. To answer your second question, soup is cheap and easily reproducible and allows for the organic variations of certain tissue types, like chicken broth for serous fluid, and green pea soup for purulent fluid. We elicited the help of a survey expert and passed it out at conferences such as this with the aim of validating our soup model. We did validate this with an inter-rater reliability using Cronbach ␣ analysis. To avoid bias, we performed the survey without background insight as to what we planned to do. We just said, “Hey, listen, could you just answer these questions? What did you think of these different types of soup in order to emulate these different kinds of bodily fluids?”