- Email: [email protected]
Conservative Management in Traumatic Pneumothoraces
[
1
Original Research
]
56
2
57
3
58
4
59
5
60
Conservative Management in Traumatic Pneumothoraces
6 7 8 9 10
Q1
61 62 63 64
An Observational Study
65
11 12 13
66 Q16
Q2 Q3
14
67
Steven P. Walker, MBChB; Shaney L. Barratt, BMBS, PhD; Julian Thompson, MD(Res);
68
and Nick A. Maskell, DM, FCCP
69
15
70
16
71
17
Traumatic pneumothoraces are a common consequence of major trauma. Despite this, there is a paucity of literature regarding their optimal management, including the role of conservative treatment. The aim of this study was to assess the treatment, complications, and outcomes of traumatic pneumothoraces in patients presenting to a major trauma center.
BACKGROUND:
18 19 20 21 22
The prospectively collected Trauma Audit and Research Network (TARN) database was used to identify all patients presenting with traumatic pneumothoraces to a UK major trauma center from April 2012 to December 2016. Demographics, mechanism of injury, injury severity score (ISS), management, and outcomes were analyzed. METHODS:
23 24 25 26 27
Six hundred two patients were included during the study period. Mean age was 48 years (SD, 22 years), and 73% were men. Mean ISS was 26 and inpatient mortality was 9%. Of the 602 traumatic pneumothoraces, 277 of 602 (46%) were initially treated conservatively. Two hundred fifty-two of 277 patients in this cohort (90%) did not require subsequent chest tube insertion, including the majority of patients (56 of 62 [90%]) who were receiving positive pressure ventilation (PPV) on admission. The hazard ratio (HR) for failure of conservative management showed no difference between the ventilated and nonventilated patients (HR, 1.1; P ¼ .84). Only the presence of a large hemothorax was associated with an increased likelihood of failure of conservative management.
RESULTS:
28 29 30 31 32 33 34 35 36 37
In the largest observational study of traumatic pneumothoraces published to date, > 90% of patients whose pneumothorax was managed conservatively never required subsequent tube drainage. Importantly, this also applies to patients requiring PPV, with no significant increased risk of failure of expectant management. These data support a role for conservative management in traumatic pneumothoraces. CHEST 2017; -(-):---
CONCLUSIONS:
38 39 40 41 42 43 44
Q8
KEY WORDS:
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
pneumothorax; trauma; ventilation
99
45
100
46
101
47
102
48 49 50 51 52 53 54 55
Q4 Q5
ABBREVIATIONS: GCS = Glasgow Coma Scale; HR = hazard ratio; ISS = Injury Severity Score; IQR = interquartile range; MTC = major trauma center; PPV = positive pressure ventilation; TARN = Trauma Audit and Research Network AFFILIATIONS: From the Academic Respiratory Unit (Drs Walker, Barratt, and Maskell), School of Clinical Sciences, University of Bristol; the North Bristol Lung Centre (Drs Walker, Barratt, and Maskell), Southmead Hospital; the Intensive Care Unit (Dr Thompson), Southmead Hospital; and the Severn Major Trauma Network (Dr Thompson), North Bristol NHS Trust, Bristol, England.
FUNDING SUPPORT:
The authors have reported to CHEST that no Q6 funding was received for this study. CORRESPONDENCE TO: Julian Thompson, MD(Res), Intensive Care Unit, Southmead Hospital, Bristol, UK; e-mail: juilian.thompson@nbt. Q7 nhs.uk Copyright Ó 2017 American College of Chest Physicians. Published by Elsevier Inc. All rights reserved. DOI: https://doi.org/10.1016/j.chest.2017.10.015
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
105 106 107 108 109 110
1
chestjournal.org
103 104
111
Traumatic pneumothoraces are present in one-fifth of patients who have undergone multiple trauma1 and is the most common potentially life-threatening injury in blunt chest trauma.2 Thoracic trauma occurs in nearly twothirds of multiple trauma cases and represents the primary cause of death in 25% of trauma patients.3 Trauma as a whole is a major public health problem, with > 150,000 deaths and > 3 million nonfatal injuries per year in the United States,4 representing the leading cause of death for individuals younger than 45 years of age.5
112 113 114 115 116 117 118 119 120 121 122
Although uncomplicated traumatic pneumothoraces may be well tolerated, the risk of tension and resultant cardiorespiratory compromise makes identification important, particularly since the basic procedure of tube thoracostomy insertion can potentially avert significant morbidity and mortality.6 Current guidance by the American College of Surgeons Advanced Trauma Life Support7 advises chest tube placement for any traumatic pneumothorax, although it suggests that asymptomatic pneumothoraces can be managed with observation and aspiration at the treating physician’s discretion. It does, however, state that a chest drain is required in patients receiving either general anesthesia or positive pressure ventilation (PPV) to avoid a life-threatening pneumothorax.7
123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138
This concept has been challenged, principally with the increased use of CT imaging in trauma that has identified small subcentimeter-sized pneumothoraces. This has raised the question of whether these small “occult” pneumothoraces, which can make up to 76% of all traumatic pneumothoraces,8 can be left untreated, particularly when positive pressure is required. Despite an early study indicating a high rate of tension in conservatively management patients receiving PPV,9 studies6,8,10-12 suggest that occult pneumothoraces can be managed expectantly, including in patients receiving PPV. Whether these results can translate to all traumatic pneumothoraces is unclear and needs clarification. CT scanning is now almost ubiquitous in patients with multiple trauma, and the distinction between occult and overt pneumothoraces may become antiquated. Case studies have also suggested that it is possible to treat larger pneumothoraces with observation.2,13 The aim of our study was to use a large trauma patient cohort to create a profile of consecutive patients presenting with traumatic pneumothoraces to examine the outcomes of conservatively managed patients and determine whether there are factors that can help predict whether a chest tube is required.
Methods
142 143 144 145 146 147 148 149 150 Q9
153 154 156 157 158 159 160 161 162 163 164 165
170 171 172 173 174 175 176 177 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 195
141
155
169
194
140
152
167 168
178
139
151
166
Q10
From April 2012 to November 2016, patients were identified as part of the Trauma Audit and Research Network (TARN) of patients presenting to the ED at Southmead Hospital, a regional UK adult major trauma center (MTC). It covers a population of 2.3 million, is supported by six other trauma unit hospitals, and admits > 1,000 major trauma patients per year. The TARN registry is a prospective observational registry of hospitalized patients who have undergone major trauma in England and Wales. TARN has given ethical approval (Section 251) for research on the anonymized data submitted by member hospitals. The TARN database includes all trauma patients irrespective of age who have a direct admission or are transferred to a member hospital and whose length of stay is 3 days or more, as well as those admitted to a high-dependency area regardless of length of stay. It also includes deaths of trauma patients occurring in the hospital, including the ED, and those transferred to other hospitals for specialist care or for an ICU/HDU bed. Certain specific injuries were excluded, including isolated neck of femur fractures or intertrochanteric/greater trochanteric fractures in persons > 65 years. International Classification of Diseases, tenth revision, codes included were S270 (traumatic pneumothorax), S270 (traumatic pneumothorax closed), S2701 (traumatic pneumothorax open), S271 (traumatic hemothorax), S2710 (traumatic hemothorax closed), S2711 (traumatic hemothorax open), S272 (traumatic hemopneumothorax), S2720 (traumatic hemopneumothorax closed), and S2721 (traumatic hemopneumothorax open). Information was collected on demographics, injury (mechanism of injury, description of injuries, Injury Severity Score [ISS]),
2 Original Research
management (type, size of drain, length of drain placement), and pneumothorax characteristics (laterality, size, and accompanying hemothorax). Airway support was characterized as either requiring PPV before hospital admission or in the ED (initial PPV) or requiring PPV subsequently due to general anesthesia administration or clinical deterioration (subsequent PPV). The size of the pneumothorax was obtained from chest radiographs at the hilum and apex, and on CT imaging, the largest collection was measured along a line perpendicular from the chest wall to the lungs or mediastinum. Clinical parameters were obtained from initial observations on attendance in the ED. Respiratory distress was determined if respiratory rate was either $ 30 or < 8, if supplementary oxygen or mechanical or manual ventilation was used, if the oxygen saturation was # 90%, or if the patient was in respiratory arrest. Hemodynamic instability was determined if the systolic BP was < 90 mm Hg or the heart rate was $ 100 bpm. Conscious-level impairment was determined if the Glasgow Coma Scale (GCS) score was < 15 or the patient was ventilated at arrival.
196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211
Statistical analysis
212
Descriptive statistics were used to summarize patient characteristics and clinical data. Means ( SD) were calculated for parametric data, and medians (interquartile range [IQR]) were calculated for nonparametric data. Several checks for normality, including Kolmogorov-Smirnov, Shapiro-Wilk, kurtosis, and skewness calculations, were performed. Continuous parametric variables were analyzed using independent t tests, and continuous nonparametric variables were analyzed using the Mann-Whitney test. Categorical data were analyzed using the c2 test. P < .05 was considered statistically significant.
[
-#- CHEST - 2017
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
]
213 214 215 216 217 218 219 220
221
Univariate proportional hazard ratios (HRs) were calculated using Cox regression analysis for factors associated with failure of conservative treatment, for example, size of pneumothorax, mechanism of injury, ISS, presence of rib fractures, clinical features (respiratory, hemodynamic, GCS score), presence of hemothorax, bilateral vs unilateral pneumothorax, use of PPV, and surgical procedures. Further multivariable Cox regression analysis was performed to
222 223 224 225 226
276
determine which factors (age, size of pneumothorax, ISS, presence of rib fractures, clinical conditions [respiratory, hemodynamic, GCS score], presence of hemothorax, bilateral vs unilateral pneumothorax, and use of PPV) were independently predictive of failure of conservative management. These factors were decided in an a priori statistical analysis plan. All statistical analysis was performed using SPSS, version 23.0 (SPSS, Inc.).
277 278 279 280 281
227
282
228 230 231
283
Results
229 Q11
Demographics
Three thousand seven hundred seventy-one trauma patients presenting to this MTC were registered in the TARN database from April 2012 to December 2016. Seven hundred sixty-five patients were identified using the search criteria. Six hundred thirty-six patients with pneumothoraces were identified, with 602 patients included for analysis (Fig 1). Table 1 summarizes patient demographics, mechanism of injuries, ISS, pneumothorax characteristics, management, and outcomes for patients managed nonconservatively and those managed conservatively. Table 2 summarizes the characteristics and outcomes for successful and failed observational management.
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246
Traumatic pneumothoraces were present in 636 of 3,771 trauma patients (17%) during the study period.
247 248
The mean age was 48 years (SD, 22 years), 438 of 602 patients (73%) were men, and 330 of 602 patients (55%) received their injuries as a result of road traffic accidents. The mean ISS score (26) represented very severe injuries, 189 of 602 patients (31%) required immediate invasive ventilation, and 56 of 602 patients (9%) died during admission. Three hundred twenty-five of 602 patients (54%) had an intervention performed before hospital admission or on admission, which included needle decompression, chest tube insertion, or chest surgery. In the remaining 277 of 602 patients (46%), the pneumothoraces were initially treated conservatively. The patients who were managed conservatively had significantly smaller pneumothoraces compared with the patients managed with immediate interventions (median, 5.5 mm vs 22 mm), with the majority of lesions < 10 mm (Fig 2).
284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
249
304
250
305
251
306
3,771 Trauma patients
252
307
253
308
254
309
3,006 excluded: • No pneumohemothorax
255
310
256
311
257
312
765 Hemopneumothoraces
258 259
313 314
163 excluded: • 89 no pneumothorax • 52 inadequate images • 20 no patient details • 2 nontraumatic pneumothoraces
260 261 262 263
315 316 317 318
602 Pneumothoraces included
264 265
319 320
266
321
267
322
268 270 271 272 273 274 275
325 Chest intervention
print & web 4C=FPO
269
323
277 Initial conservative management
324 325 326 327
252 Successful conservative management
25 Failed conservative management
328 329
Figure 1 – Cohort diagram. Demonstration of the numbers of patients included in the study, providing reasons for noninclusion where necessary.
Q17
3
chestjournal.org
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
330
331
TABLE 1
] Characteristics and Outcomes in Patients Receiving Nonconservative vs Conservative Management
332
Initial Nonconservative Management (n ¼ 325)
Conservative Management (n ¼ 277)
Age, SD, y
48.8 (21.5)
47.1 (21.9)
336
Male sex, %
252 (77.5)
186 (67.1)
337
Mechanism of injury, %
333 334 335
Q15
Variable
386 387 388
P Value
389
.36
390
.04
391
.04
392
338
Vehicle collision
175 (53.8)
155 (56)
393
339
Fall < 2 m
58 (17.8)
55 (19.9)
394
340
Fall > 2 m
52 (16)
48 (17.3)
395
25 (7.7)
5 (1.8)
396
7 (2.2)
8 (2.9)
341
Stabbings
342
Crush injuries
343 344
Blows
397 398
8 (2.5)
6 (2.2)
26.9 (14.7)
25.0 (12.3)
.08
399
345
Mean ISS, SD
346
Median pneumothorax size, mm, IQR
22 (35.9)
5.5 (8.8)
< .001
347
Significant (> 2 cm) hemothorax, %
58 (17.8)
17 (6.1)
< .001
348
Initial positive pressure ventilation, %
127 (39.1)
62 (22.4)
< .001
403 404
400 401 402
349
Subsequent positive pressure ventilation, %
159 (48.9)
139 (50.2)
.12
350
Presence of respiratory distress, %
199 (61.2)
138 (49.8)
.06
Presence of hemodynamic compromise, %
132 (40.6)
71 (25.6)
< .001
353
Presence of decreased GCS score, %
135 (41.5)
90 (32.5)
.03
408
354
Median hospital length of stay, IQR, d
10 (17)
10 (13.5)
.35
409
355
Median ICU length of stay, IQR, d
2 (9.5)
0 (5)
< .001
410
356
Mortality, %
11.1%
7.2%
.1
351 352
357 358
405 406 407
411 412
GCS ¼ Glasgow Coma Scale; IQR ¼ interquartile range.
413
359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385
414
Patients who were managed with immediate interventions also had a higher incidence of respiratory, hemodynamic, and neurologic compromise and a higher proportion of significant hemothoraces than did those managed conservatively. Both groups had comparable ages, ISS data, mortality rates, and total length of stay. Of the 277 of 602 patients (46%) managed conservatively, 252 of 277 (90%) did not require subsequent thoracic intervention. This included the majority of patients (56 of 62 [90%]) requiring immediate PPV who were treated conservatively. There was no significant difference in the failure rate between the patients receiving PPV (6 of 62 [9.7%]) and those not requiring PPV (19 of 215 [8.8%]) in the conservative arm. Using univariate analysis, the size of the pneumothorax, mechanism of injury, presence of rib fractures, clinical condition, surgery, and ISS were not significantly associated with failure of conservative management (Table 3). The median size of the pneumothorax (5.3 mm vs 8.2 mm; P ¼ .13) was comparable between groups and did not increase the likelihood of progression requiring chest tube insertion, with HRs of
4 Original Research
415
1.61 (P ¼ .08) and 2.84 (P ¼ .07) on univariate and multivariable analyses, respectively.
416 417
Univariate and multivariable analyses also confirmed that acute PPV does not appear to confer an additional risk of failure of conservative management (HR, 1.1; P ¼ .8 and HR, 1.5; P ¼ .96, respectively). Additionally, requiring subsequent PPV during an inpatient stay due to clinical deterioration or for general anesthesia did not represent an increased risk of failure of conservative management. In contrast, the presence of a hemothorax was associated with an increased likelihood of failure of expectant management (HR, 4.08; P < .01), which was confirmed by multivariable Cox regression analysis (Table 4). Of the 25 of 252 patients in whom observational management failed, 23 had a large chest drain inserted and two went on to have thoracic surgery (rib fixation with hemothorax evacuation). The main indication for chest tube insertion was increasing pneumothorax (19 of 23 cases) and enlarging hemopneumothorax (4 of 23 cases). The mean duration prior to chest tube insertion was 2.96 days (SD, 4.03 days). Requiring subsequent chest tube insertion in the conservative arm led to a
[
-#- CHEST - 2017
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
]
418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
441
TABLE 2
] Characteristics and Outcomes of Conservatively Managed Patients
442
496 497
Conservative Management (Successful) (n ¼ 252)
Conservative Management (Failed) (n ¼ 25)
P Value
Age, SD, y
46.7 (22.4)
51.2 (16.2)
.33
446
Male sex, %
169 (67.1)
17 (68)
.92
447
Mechanism of injury, %
443
Variable
444 445
498 499 500 501
.72
502
448
Vehicle collision
142 (56.3)
13 (53)
503
449
Fall < 2 m
49 (19.4)
6 (24)
504
450
Fall > 2 m
42 (16.7)
6 (24)
505
5 (2)
0
506
8 (3.2)
0
451
Stabbings
452
Crush injuries
453
Blows
454
507 508
6 (2.4)
0
24.9 (12.5)
25.0 (11.0)
.97 .13
511 512
509
455
Mean ISS, SD
456
Median pneumothorax size, IQR, mm
5.3 (8.6)
8.2 (16.5)
457
Significant (> 2 cm) hemothorax, %
12 (4.9)
5 (20)
< .01
458
Initial positive pressure ventilation, %
56 (22.2)
6 (24)
.84
513 514
459
Subsequent positive pressure ventilation, %
460
Presence of Respiratory distress, %
510
123 (49)
16 (64)
.15
124 (50.2)
14 (60.9)
.33
Presence of hemodynamic compromise, %
67 (27.2)
4 (18.2)
.36
463
Presence of decreased GCS, %
80 (32.3)
10 (40)
.43
464
Median hospital length of stay, IQR, d
10 (13.8)
11 (14.5)
.66
519
465
Median ICU length of stay, IQR, d
0 (4)
3 (12.0)
.15
520
466
Mortality, %
7.1
8
.88
461 462
467
515 516 517 518
521 522
ISS ¼ Injury Severity Score. See Table 1 legend for expansion of other abbreviations.
468
523
469
524
470
525
nonsignificant increased length of stay (11 vs 10 days; P ¼ .597). The two patients with mortalities in this group had an ISS score of 40 (> 25 represents severe or critical injuries) with intracranial hemorrhage, and it is unlikely that the pneumothorax contributed to the overall outcome. The rate of cardiothoracic surgery was
471 472 473 474 475 476 477 478 479 480
100.0
481 482
80.0
483
Frequency
484 485 486 488 489 490 491 492 493 494 495
print & web 4C=FPO
487
60.0 40.0 20.0 0.0 .00
10.00 20.00 30.00 40.00 50.00 60.00 Pneumothorax Size (mm)
Figure 2 – Pneumothorax size in conservatively managed patients.
higher in the intervention cohort, with 18 of 325 patients requiring surgery. Table 5 shows the characteristics of patients with a pneumothorax visible on chest radiography (overt pneumothorax) and of patients with a pneumothorax not visible on chest radiography (occult pneumothorax). One hundred seventy-seven patients had chest radiography as their initial chest imaging. One hundred thirty-seven of these patients proceeded to chest CT imaging. Of these 137 patients, 11 had a chest drain in situ at the time of chest radiography. Of the remaining 126 patients, 61 of 126 (48%) had no visible pneumothorax. Occult pneumothoraces were generally smaller than the overt pneumothoraces, with a respective median size of 7.26 mm vs 25.07 mm (P < .001). The majority, 470 of 603 study patients (78%), had evidence of rib fractures, with 361 of 470 patients and 143 of 470 patients having more than three and five rib fractures, respectively. Four hundred twenty-seven of 470 of the fractures were unilateral, and 167 of 470 were reported as flail chest, with 12 of 470 reported as
5
chestjournal.org
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
551 552
TABLE 3
] Hazard Ratios for Failed Conservative Management
606 607
Hazard Ratio
P Value
Male sex
1.05
.92
0.45-2.2
555
Size of pneumothorax ($ 2 cm vs < 2 cm)
1.61
.08
0.94-2.76
556
Bilateral vs unilateral pneumothorax
1.34
.25
0.83-2.12
611
557
ISS score (very severe vs severe and moderate severe)
1.17
.69
0.54-2.58
612
558
Presence of rib fractures
1.15
.57
0.71-1.88
613 614
553 554
559 560 561
Variable
Hemothorax (> 2 cm) Received initial positive pressure ventilation
95% CI
4.08
< .01
1.53-10.88
1.1
.84
0.44-2.76
608 609 610
615 616
562
Received subsequent positive pressure ventilation
2.10
.08
0.91-4.87
563
Presence of respiratory distress
1.23
.33
0.810-1.87
564
Presence of hemodynamic compromise
0.78
.37
0.45-1.34
619
565
Presence of decreased GCS score
1.17
.45
0.78-1.74
620
566 567
617 618
621
See Table 1 and 2 legends for expansion of abbreviations.
622
568 569 570 571 572 573 574 575 576 577 578 579 580 581 582
623
bilateral flail chest. There was no statistically significant association between the presence of rib fractures and significant hemothorax (c2 ¼ 0.946; P ¼ .331). There was a 10% complication rate associated with chest tube insertion. Fifteen patients (4.4%) required their drains to be resited, four patients (1.2%) had their drains dislodged, five patients (1.5%) had intraparenchymal drains on CT imaging, two patients (0.6%) experienced empyema, and one patient (0.3%) had a guidewire left in the pleural cavity. Eight patients (2.4%) required a subsequent drain after initial removal due to reaccumulation of air or fluid.
583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598
Discussion Chest drain insertion is not without risk of complications, with documented complication rates ranging from 15% to 30%.12,14-16 Current guidelines recommend chest tube placement for traumatic pneumothorax, particularly in patients receiving PPV,7 with a caveat that asymptomatic nonventilated patients can be managed with observation or aspiration at the treating clinician’s discretion. Existing literature has examined whether occult pneumothoraces can be managed conservatively.6,12 Scoring systems to determine whether chest tube intervention is required for occult pneumothoraces are in their infancy and have
599 600 601 602 603 604 605
TABLE 4
] Multivariable Cox Regression Analysis for Failure of Conservative Management
Variable Hemothorax > 2 cm
Hazard Ratio
P Value
5.29
< .01
95% CI 1.78-15.79
not been prospectively validated.17 We sought to determine whether traumatic pneumothoraces can be treated conservatively and examined factors that safely identify patients who could avoid chest tube insertion. In this paper, we show that the majority of patients managed conservatively, including patients receiving PPV, did not require further invasive ventilation. Studies have focused on whether there is a role for conservative management for occult pneumothoraces that are not initially visible on chest radiography.6,12 The resultant positive findings, including in patients receiving PPV, have been incorporated into clinical guidelines.18 Although these studies have been useful in establishing management pathways for traumatic pneumothoraces, they do have limitations. It is difficult to translate their findings into clinical practice in which CT is becoming the first-line investigation, with the majority of patients (70%) in our study having had a CT scan as their initial imaging and > 90% undergoing CT imaging during the inpatient stay. Furthermore, the distinction between overt and occult can be misleading. TABLE 5
Variable Age, SD, y
52.04 (23.80)
Overt Pneumothoraces (n ¼ 65) 56.30 (20.45)
631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647
48 (78.7) 7.26 (12.39)
651 652 653 654
47 (72.3)
655 656
25.07 (37.46)
657 658 20.00 (9.50)
16.00 (11)
659 660
See Table 2 legend for expansion of abbreviations.
6 Original Research
629 630
650
Occult Pneumothoraces (n ¼ 61)
Median ISS, IQR
627 628
649
Characteristics
Median size of pneumothorax on CT, IQR, mm
625 626
648
] Overt and Occult Pneumothoraces
Male sex, %
624
[
-#- CHEST - 2017
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
]
661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715
Whether a pneumothorax is seen on a chest radiograph is not solely related to its size and can be influenced by other factors, for example, the use of supine chest radiography has decreased sensitivity,8 with occult pneumothoraces in our study reaching > 80 mm in size. With this is mind, we proposed to look at the outcomes for traumatic pneumothoraces as a whole. Nearly half of the patients (46%) included were managed conservatively, with the majority of them (252 of 277 [90%]) not requiring subsequent invasive treatment for their pneumothoraces. This included 56 of 62 patients (90%) who received immediate PPV. Multivariable analysis supported the finding that immediate or subsequent PPV did not confer an additional risk of failure of conservative management. This is consistent with the most current study on occult pneumothoraces, with Moore et al12 demonstrating a failure rate of 14% in patients receiving PPV managed with observation. Although this was higher than their 4.5% failure rate for those not receiving PPV, the latter was not identified as an independent predictor of failed management on multivariate analysis, and no patient experienced tension pneumothorax related to delayed tube insertion. Smaller earlier studies found conflicting results. Brasel et al11 and Enderson et al9 demonstrated a 22% and 53% failure rate, respectively, for patients with occult pneumothoraces who were receiving PPV and were managed with observation. Brasel et al’s11 paper concluded that observation was safe in these patients, whereas Enderson et al9 recommended tube thoracostomy for all patients requiring PPV. This has led to an ongoing debate regarding the management of occult pneumothorax in patients receiving PPV. The East Practice Management Guidelines (2011) on occult pneumothoraces18 recommend that occult pneumothoraces may be observed in stable patients regardless of PPV. The currently recruiting Management of Occult Pneumothoraces in Mechanically Ventilated Patients (OPTICC) trial (ClinicalTrials.gov: NCT00530725), which is randomizing occult pneumothoraces in patients receiving PPV between chest tube insertion and observation, should contribute to this evidence base. The size of the pneumothorax was not a predictor of failed observation in our study on univariate and multivariate analyses, with nonsignificant differences in the size of pneumothoraces between the successful and failed observed groups. Pneumothorax size had previously been thought to be a predictor of progression,19 with De Moya et al17 proposing a scoring
system using the size of the occult pneumothorax and its relationship to the hilum to guide management. However, this has not been successfully validated,12 with Moore et al12 demonstrating that pneumothorax size was not an independent predictor of failed observation. The presence of a hemothorax appears to be predictive of failure of conservative management in both overt and occult pneumothoraces.12 This is consistent with clinical practice. A significant hemothorax is an indication for chest tube insertion to evacuate blood from the pleural space and avoid complications such as infection and fibrothorax, and when combined with the presence of a pneumothorax, this provides a strong incentive for intervention. In this study, patients who did not require prehospital or admission chest procedures generally did not require a chest procedure later during hospitalization. When this information is combined with previous trials on traumatic pneumothoraces, it appears that there is a subpopulation that can be managed conservatively. Certainly, when there is no significant hemopneumothorax (< 2 cm), there can be consideration for expectant management. Mechanism of injury, ISS, or size of the pneumothorax does not appear to provide a strong indication for intervention. Additionally, in our study, clinical condition did not confer an adverse prognosis, although in other studies, respiratory distress did.12 Although the use of ventilation has been controversial, it appears from our findings and previous studies that pneumothoraces can be managed conservatively with careful observation in patients receiving PPV with no increased risk of harm.6,12
716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752
This is a retrospective observational trial and as such is subject to the inherent limitations of such a study. The data were collected from a single center with a low rate of penetrating chest wall injuries (5%), which should be considered when generalizing the findings to other centers. Selection bias may have been introduced by physician selection and variation in initial imaging modality, and the decision to intervene may have affected the conservatively treated cohort characteristics, and it is likely that high-risk unwell patients were underrepresented in the conservatively treated arm. Those treated with an immediate intervention, despite a similar ISS, likely represented a more unwell population, with higher rates of cardiorespiratory compromise, PPV use, surgical referral rates, and mortality. The length of stay criteria (length of stay ¼ 3 days or more, or admittance to a high-dependency area regardless of
7
chestjournal.org
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770
771
length of stay) is likely to have been biased against patients conservatively managed successfully and not requiring a prolonged hospital admission, suggesting that the overall rate of effective conservative management is probably greater. Efforts were made to minimize bias by including a large number of consecutive unselected patients into the analysis and careful documentation and comparison of cohort characteristics.
772 773 774 775 776 777 778 779 780 781 782
Conclusions
783
Our study represents the largest observational study on traumatic pneumothoraces to date. It demonstrates that
784
the majority of conservatively managed patients were successfully managed without requiring a chest drain. This includes the majority of patients receiving PPV, the use of which did not present an increased risk of failure of expectant management. This study provides support for an observational expectant approach if the treating physician does not believe that an immediate chest drain is warranted in a patient with a traumatic pneumothorax. Future prospective randomized trials examining the outcomes of a conservative approach in traumatic pneumothorax, regardless of pneumothorax size or use of PPV, would help clarify which patients are best managed expectantly.
785
826 827 828 829 830 831 832 833 834 835 836 837 838 839 840
786
841
787
842
788
Acknowledgments
789
Author contributions: S. P. W. takes responsibility for the content of the manuscript, including the data and analysis. S. P. W. performed statistical analysis and prepared the manuscript. S. P. W., S. L. B., J. T., and N. A. M. conceived the design of the study. All authors read and approved the final manuscript for submission.
790 791
Q14
792 793 794 795
Q12
796 797 798 799
References
800
1. Di Bartolomeo S, Sanson G, Nardi G, Scian F, Michelutto V, Lattuada L. A population-based study on pneumothorax in severely traumatized patients. J Trauma Acute Care Surg. 2001;51(4):677-682.
801 802 803 804 805
2. Idris BM, Hefny AF. Large pneumothorax in blunt chest trauma: is a chest drain always necessary in stable patients? A case report. Int J Surg Case Rep. 2016;24:88-90.
806 807 808
3. Kshettry VR, Bolman R, 3rd. Chest trauma. Assessment, diagnosis, and management. Clin Chest Med. 1994;15(1):137.
809 810 811 812 813 814
Financial/nonfinancial disclosures S. L. B. has received funding for accommodation and registration fees from Boehringer Ingelheim. None declared (S. P. W., J. T., N. M.).
Q13
4. Centers for Disease Control and Prevention NCfHSN. National hospital discharge survey: 2007 summary. National health statistics reports, no. 29. National health statistics reports, no. 29. Atlanta, GA: National Center for Health Statistics; 2010.
5. Centers for Disease Control and Prevention. Web-based Injury Statistics Query and Reporting System (WISQARS), 2007. http://www.cdc.gov/ injury/wisqars. Accessed xxx. 6. Wilson H, Ellsmere J, Tallon J, Kirkpatrick A. Occult pneumothorax in the blunt trauma patient: tube thoracostomy or observation? Injury. 2009;40(9):928-931. 7. ATLS Subcommittee; American College of Surgeons’ Committee on Trauma; International ATLS working group. Advanced trauma life support (ATLSÒ): the ninth edition. J Trauma Acute Care Surg. 2013;74(5):1363. 8. Ball CG, Kirkpatrick AW, Feliciano DV. The occult pneumothorax: what have we learned? Can J Surg. 2009;52(5):E173. 9. Enderson BL, Abdalla R, Frame SB, Casey MT, Gould H, Maull KI. Tube thoracostomy for occult pneumothorax: a prospective randomized study of its use. J Trauma Acute Care Surg. 1993;35(5): 726-730. 10. Plurad D, Green D, Demetriades D, Rhee P. The increasing use of chest computed tomography for trauma: is it being overutilized? J Trauma Acute Care Surg. 2007;62(3):631-635. 11. Brasel KJ, Stafford RE, Weigelt JA, Tenquist JE, Borgstrom DC. Treatment of occult pneumothoraces from blunt trauma. J Trauma Acute Care Surg. 1999;46(6):987-991.
12. Moore FO, Goslar PW, Coimbra R, et al. Blunt traumatic occult pneumothorax: is observation safe?—results of a prospective, AAST multicenter study. J Trauma Acute Care Surg. 2011;70(5): 1019-1025. 13. Ryan MT, Caputo ND, Lakdawala V, Jara F. Spontaneous resolution of a large traumatic pneumothorax. Am J Emerg Med. 2012;30(5):833.e3. 14. Etoch SW, Bar-Natan MF, Miller FB, Richardson JD. Tube thoracostomy: factors related to complications. Arch Surg. 1995;130(5):521-526.
843 844 845 846 847 848 849 850 851 852 853
15. Bailey R. Complications of tube thoracostomy in trauma. J Accid Emerg Med. 2000;17(2):111-114.
854
16. Ball CG, Lord J, Laupland KB, et al. Chest tube complications: how well are we training our residents? Can J Surg. 2007;50(6):450.
856
855
17. De Moya MA, Seaver C, Spaniolas K, et al. Occult pneumothorax in trauma patients: development of an objective scoring system. J Trauma Acute Care Surg. 2007;63(1):13-17. 18. Mowery NT, Gunter OL, Collier BR, et al. Practice management guidelines for management of hemothorax and occult pneumothorax. J Trauma Acute Care Surg. 2011;70(2):510-518. 19. Garramone R, Jr, Jacobs L, Sahdev P. An objective method to measure and manage occult pneumothorax. Surg Gynecol Obstet. 1991;173(4):257-261.
815
857 858 859 860 861 862 863 864 865 866 867 868 869 870
816
871
817
872
818
873
819
874
820
875
821
876
822
877
823
878
824
879
825
880
8 Original Research
[
-#- CHEST - 2017
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
]
1
Original Research
]
56
2
57
3
58
4
59
5
60
Conservative Management in Traumatic Pneumothoraces
6 7 8 9 10
Q1
61 62 63 64
An Observational Study
65
11 12 13
66 Q16
Q2 Q3
14
67
Steven P. Walker, MBChB; Shaney L. Barratt, BMBS, PhD; Julian Thompson, MD(Res);
68
and Nick A. Maskell, DM, FCCP
69
15
70
16
71
17
Traumatic pneumothoraces are a common consequence of major trauma. Despite this, there is a paucity of literature regarding their optimal management, including the role of conservative treatment. The aim of this study was to assess the treatment, complications, and outcomes of traumatic pneumothoraces in patients presenting to a major trauma center.
BACKGROUND:
18 19 20 21 22
The prospectively collected Trauma Audit and Research Network (TARN) database was used to identify all patients presenting with traumatic pneumothoraces to a UK major trauma center from April 2012 to December 2016. Demographics, mechanism of injury, injury severity score (ISS), management, and outcomes were analyzed. METHODS:
23 24 25 26 27
Six hundred two patients were included during the study period. Mean age was 48 years (SD, 22 years), and 73% were men. Mean ISS was 26 and inpatient mortality was 9%. Of the 602 traumatic pneumothoraces, 277 of 602 (46%) were initially treated conservatively. Two hundred fifty-two of 277 patients in this cohort (90%) did not require subsequent chest tube insertion, including the majority of patients (56 of 62 [90%]) who were receiving positive pressure ventilation (PPV) on admission. The hazard ratio (HR) for failure of conservative management showed no difference between the ventilated and nonventilated patients (HR, 1.1; P ¼ .84). Only the presence of a large hemothorax was associated with an increased likelihood of failure of conservative management.
RESULTS:
28 29 30 31 32 33 34 35 36 37
In the largest observational study of traumatic pneumothoraces published to date, > 90% of patients whose pneumothorax was managed conservatively never required subsequent tube drainage. Importantly, this also applies to patients requiring PPV, with no significant increased risk of failure of expectant management. These data support a role for conservative management in traumatic pneumothoraces. CHEST 2017; -(-):---
CONCLUSIONS:
38 39 40 41 42 43 44
Q8
KEY WORDS:
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
pneumothorax; trauma; ventilation
99
45
100
46
101
47
102
48 49 50 51 52 53 54 55
Q4 Q5
ABBREVIATIONS: GCS = Glasgow Coma Scale; HR = hazard ratio; ISS = Injury Severity Score; IQR = interquartile range; MTC = major trauma center; PPV = positive pressure ventilation; TARN = Trauma Audit and Research Network AFFILIATIONS: From the Academic Respiratory Unit (Drs Walker, Barratt, and Maskell), School of Clinical Sciences, University of Bristol; the North Bristol Lung Centre (Drs Walker, Barratt, and Maskell), Southmead Hospital; the Intensive Care Unit (Dr Thompson), Southmead Hospital; and the Severn Major Trauma Network (Dr Thompson), North Bristol NHS Trust, Bristol, England.
FUNDING SUPPORT:
The authors have reported to CHEST that no Q6 funding was received for this study. CORRESPONDENCE TO: Julian Thompson, MD(Res), Intensive Care Unit, Southmead Hospital, Bristol, UK; e-mail: juilian.thompson@nbt. Q7 nhs.uk Copyright Ó 2017 American College of Chest Physicians. Published by Elsevier Inc. All rights reserved. DOI: https://doi.org/10.1016/j.chest.2017.10.015
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
105 106 107 108 109 110
1
chestjournal.org
103 104
111
Traumatic pneumothoraces are present in one-fifth of patients who have undergone multiple trauma1 and is the most common potentially life-threatening injury in blunt chest trauma.2 Thoracic trauma occurs in nearly twothirds of multiple trauma cases and represents the primary cause of death in 25% of trauma patients.3 Trauma as a whole is a major public health problem, with > 150,000 deaths and > 3 million nonfatal injuries per year in the United States,4 representing the leading cause of death for individuals younger than 45 years of age.5
112 113 114 115 116 117 118 119 120 121 122
Although uncomplicated traumatic pneumothoraces may be well tolerated, the risk of tension and resultant cardiorespiratory compromise makes identification important, particularly since the basic procedure of tube thoracostomy insertion can potentially avert significant morbidity and mortality.6 Current guidance by the American College of Surgeons Advanced Trauma Life Support7 advises chest tube placement for any traumatic pneumothorax, although it suggests that asymptomatic pneumothoraces can be managed with observation and aspiration at the treating physician’s discretion. It does, however, state that a chest drain is required in patients receiving either general anesthesia or positive pressure ventilation (PPV) to avoid a life-threatening pneumothorax.7
123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138
This concept has been challenged, principally with the increased use of CT imaging in trauma that has identified small subcentimeter-sized pneumothoraces. This has raised the question of whether these small “occult” pneumothoraces, which can make up to 76% of all traumatic pneumothoraces,8 can be left untreated, particularly when positive pressure is required. Despite an early study indicating a high rate of tension in conservatively management patients receiving PPV,9 studies6,8,10-12 suggest that occult pneumothoraces can be managed expectantly, including in patients receiving PPV. Whether these results can translate to all traumatic pneumothoraces is unclear and needs clarification. CT scanning is now almost ubiquitous in patients with multiple trauma, and the distinction between occult and overt pneumothoraces may become antiquated. Case studies have also suggested that it is possible to treat larger pneumothoraces with observation.2,13 The aim of our study was to use a large trauma patient cohort to create a profile of consecutive patients presenting with traumatic pneumothoraces to examine the outcomes of conservatively managed patients and determine whether there are factors that can help predict whether a chest tube is required.
Methods
142 143 144 145 146 147 148 149 150 Q9
153 154 156 157 158 159 160 161 162 163 164 165
170 171 172 173 174 175 176 177 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 195
141
155
169
194
140
152
167 168
178
139
151
166
Q10
From April 2012 to November 2016, patients were identified as part of the Trauma Audit and Research Network (TARN) of patients presenting to the ED at Southmead Hospital, a regional UK adult major trauma center (MTC). It covers a population of 2.3 million, is supported by six other trauma unit hospitals, and admits > 1,000 major trauma patients per year. The TARN registry is a prospective observational registry of hospitalized patients who have undergone major trauma in England and Wales. TARN has given ethical approval (Section 251) for research on the anonymized data submitted by member hospitals. The TARN database includes all trauma patients irrespective of age who have a direct admission or are transferred to a member hospital and whose length of stay is 3 days or more, as well as those admitted to a high-dependency area regardless of length of stay. It also includes deaths of trauma patients occurring in the hospital, including the ED, and those transferred to other hospitals for specialist care or for an ICU/HDU bed. Certain specific injuries were excluded, including isolated neck of femur fractures or intertrochanteric/greater trochanteric fractures in persons > 65 years. International Classification of Diseases, tenth revision, codes included were S270 (traumatic pneumothorax), S270 (traumatic pneumothorax closed), S2701 (traumatic pneumothorax open), S271 (traumatic hemothorax), S2710 (traumatic hemothorax closed), S2711 (traumatic hemothorax open), S272 (traumatic hemopneumothorax), S2720 (traumatic hemopneumothorax closed), and S2721 (traumatic hemopneumothorax open). Information was collected on demographics, injury (mechanism of injury, description of injuries, Injury Severity Score [ISS]),
2 Original Research
management (type, size of drain, length of drain placement), and pneumothorax characteristics (laterality, size, and accompanying hemothorax). Airway support was characterized as either requiring PPV before hospital admission or in the ED (initial PPV) or requiring PPV subsequently due to general anesthesia administration or clinical deterioration (subsequent PPV). The size of the pneumothorax was obtained from chest radiographs at the hilum and apex, and on CT imaging, the largest collection was measured along a line perpendicular from the chest wall to the lungs or mediastinum. Clinical parameters were obtained from initial observations on attendance in the ED. Respiratory distress was determined if respiratory rate was either $ 30 or < 8, if supplementary oxygen or mechanical or manual ventilation was used, if the oxygen saturation was # 90%, or if the patient was in respiratory arrest. Hemodynamic instability was determined if the systolic BP was < 90 mm Hg or the heart rate was $ 100 bpm. Conscious-level impairment was determined if the Glasgow Coma Scale (GCS) score was < 15 or the patient was ventilated at arrival.
196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211
Statistical analysis
212
Descriptive statistics were used to summarize patient characteristics and clinical data. Means ( SD) were calculated for parametric data, and medians (interquartile range [IQR]) were calculated for nonparametric data. Several checks for normality, including Kolmogorov-Smirnov, Shapiro-Wilk, kurtosis, and skewness calculations, were performed. Continuous parametric variables were analyzed using independent t tests, and continuous nonparametric variables were analyzed using the Mann-Whitney test. Categorical data were analyzed using the c2 test. P < .05 was considered statistically significant.
[
-#- CHEST - 2017
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
]
213 214 215 216 217 218 219 220
221
Univariate proportional hazard ratios (HRs) were calculated using Cox regression analysis for factors associated with failure of conservative treatment, for example, size of pneumothorax, mechanism of injury, ISS, presence of rib fractures, clinical features (respiratory, hemodynamic, GCS score), presence of hemothorax, bilateral vs unilateral pneumothorax, use of PPV, and surgical procedures. Further multivariable Cox regression analysis was performed to
222 223 224 225 226
276
determine which factors (age, size of pneumothorax, ISS, presence of rib fractures, clinical conditions [respiratory, hemodynamic, GCS score], presence of hemothorax, bilateral vs unilateral pneumothorax, and use of PPV) were independently predictive of failure of conservative management. These factors were decided in an a priori statistical analysis plan. All statistical analysis was performed using SPSS, version 23.0 (SPSS, Inc.).
277 278 279 280 281
227
282
228 230 231
283
Results
229 Q11
Demographics
Three thousand seven hundred seventy-one trauma patients presenting to this MTC were registered in the TARN database from April 2012 to December 2016. Seven hundred sixty-five patients were identified using the search criteria. Six hundred thirty-six patients with pneumothoraces were identified, with 602 patients included for analysis (Fig 1). Table 1 summarizes patient demographics, mechanism of injuries, ISS, pneumothorax characteristics, management, and outcomes for patients managed nonconservatively and those managed conservatively. Table 2 summarizes the characteristics and outcomes for successful and failed observational management.
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246
Traumatic pneumothoraces were present in 636 of 3,771 trauma patients (17%) during the study period.
247 248
The mean age was 48 years (SD, 22 years), 438 of 602 patients (73%) were men, and 330 of 602 patients (55%) received their injuries as a result of road traffic accidents. The mean ISS score (26) represented very severe injuries, 189 of 602 patients (31%) required immediate invasive ventilation, and 56 of 602 patients (9%) died during admission. Three hundred twenty-five of 602 patients (54%) had an intervention performed before hospital admission or on admission, which included needle decompression, chest tube insertion, or chest surgery. In the remaining 277 of 602 patients (46%), the pneumothoraces were initially treated conservatively. The patients who were managed conservatively had significantly smaller pneumothoraces compared with the patients managed with immediate interventions (median, 5.5 mm vs 22 mm), with the majority of lesions < 10 mm (Fig 2).
284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
249
304
250
305
251
306
3,771 Trauma patients
252
307
253
308
254
309
3,006 excluded: • No pneumohemothorax
255
310
256
311
257
312
765 Hemopneumothoraces
258 259
313 314
163 excluded: • 89 no pneumothorax • 52 inadequate images • 20 no patient details • 2 nontraumatic pneumothoraces
260 261 262 263
315 316 317 318
602 Pneumothoraces included
264 265
319 320
266
321
267
322
268 270 271 272 273 274 275
325 Chest intervention
print & web 4C=FPO
269
323
277 Initial conservative management
324 325 326 327
252 Successful conservative management
25 Failed conservative management
328 329
Figure 1 – Cohort diagram. Demonstration of the numbers of patients included in the study, providing reasons for noninclusion where necessary.
Q17
3
chestjournal.org
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
330
331
TABLE 1
] Characteristics and Outcomes in Patients Receiving Nonconservative vs Conservative Management
332
Initial Nonconservative Management (n ¼ 325)
Conservative Management (n ¼ 277)
Age, SD, y
48.8 (21.5)
47.1 (21.9)
336
Male sex, %
252 (77.5)
186 (67.1)
337
Mechanism of injury, %
333 334 335
Q15
Variable
386 387 388
P Value
389
.36
390
.04
391
.04
392
338
Vehicle collision
175 (53.8)
155 (56)
393
339
Fall < 2 m
58 (17.8)
55 (19.9)
394
340
Fall > 2 m
52 (16)
48 (17.3)
395
25 (7.7)
5 (1.8)
396
7 (2.2)
8 (2.9)
341
Stabbings
342
Crush injuries
343 344
Blows
397 398
8 (2.5)
6 (2.2)
26.9 (14.7)
25.0 (12.3)
.08
399
345
Mean ISS, SD
346
Median pneumothorax size, mm, IQR
22 (35.9)
5.5 (8.8)
< .001
347
Significant (> 2 cm) hemothorax, %
58 (17.8)
17 (6.1)
< .001
348
Initial positive pressure ventilation, %
127 (39.1)
62 (22.4)
< .001
403 404
400 401 402
349
Subsequent positive pressure ventilation, %
159 (48.9)
139 (50.2)
.12
350
Presence of respiratory distress, %
199 (61.2)
138 (49.8)
.06
Presence of hemodynamic compromise, %
132 (40.6)
71 (25.6)
< .001
353
Presence of decreased GCS score, %
135 (41.5)
90 (32.5)
.03
408
354
Median hospital length of stay, IQR, d
10 (17)
10 (13.5)
.35
409
355
Median ICU length of stay, IQR, d
2 (9.5)
0 (5)
< .001
410
356
Mortality, %
11.1%
7.2%
.1
351 352
357 358
405 406 407
411 412
GCS ¼ Glasgow Coma Scale; IQR ¼ interquartile range.
413
359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385
414
Patients who were managed with immediate interventions also had a higher incidence of respiratory, hemodynamic, and neurologic compromise and a higher proportion of significant hemothoraces than did those managed conservatively. Both groups had comparable ages, ISS data, mortality rates, and total length of stay. Of the 277 of 602 patients (46%) managed conservatively, 252 of 277 (90%) did not require subsequent thoracic intervention. This included the majority of patients (56 of 62 [90%]) requiring immediate PPV who were treated conservatively. There was no significant difference in the failure rate between the patients receiving PPV (6 of 62 [9.7%]) and those not requiring PPV (19 of 215 [8.8%]) in the conservative arm. Using univariate analysis, the size of the pneumothorax, mechanism of injury, presence of rib fractures, clinical condition, surgery, and ISS were not significantly associated with failure of conservative management (Table 3). The median size of the pneumothorax (5.3 mm vs 8.2 mm; P ¼ .13) was comparable between groups and did not increase the likelihood of progression requiring chest tube insertion, with HRs of
4 Original Research
415
1.61 (P ¼ .08) and 2.84 (P ¼ .07) on univariate and multivariable analyses, respectively.
416 417
Univariate and multivariable analyses also confirmed that acute PPV does not appear to confer an additional risk of failure of conservative management (HR, 1.1; P ¼ .8 and HR, 1.5; P ¼ .96, respectively). Additionally, requiring subsequent PPV during an inpatient stay due to clinical deterioration or for general anesthesia did not represent an increased risk of failure of conservative management. In contrast, the presence of a hemothorax was associated with an increased likelihood of failure of expectant management (HR, 4.08; P < .01), which was confirmed by multivariable Cox regression analysis (Table 4). Of the 25 of 252 patients in whom observational management failed, 23 had a large chest drain inserted and two went on to have thoracic surgery (rib fixation with hemothorax evacuation). The main indication for chest tube insertion was increasing pneumothorax (19 of 23 cases) and enlarging hemopneumothorax (4 of 23 cases). The mean duration prior to chest tube insertion was 2.96 days (SD, 4.03 days). Requiring subsequent chest tube insertion in the conservative arm led to a
[
-#- CHEST - 2017
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
]
418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
441
TABLE 2
] Characteristics and Outcomes of Conservatively Managed Patients
442
496 497
Conservative Management (Successful) (n ¼ 252)
Conservative Management (Failed) (n ¼ 25)
P Value
Age, SD, y
46.7 (22.4)
51.2 (16.2)
.33
446
Male sex, %
169 (67.1)
17 (68)
.92
447
Mechanism of injury, %
443
Variable
444 445
498 499 500 501
.72
502
448
Vehicle collision
142 (56.3)
13 (53)
503
449
Fall < 2 m
49 (19.4)
6 (24)
504
450
Fall > 2 m
42 (16.7)
6 (24)
505
5 (2)
0
506
8 (3.2)
0
451
Stabbings
452
Crush injuries
453
Blows
454
507 508
6 (2.4)
0
24.9 (12.5)
25.0 (11.0)
.97 .13
511 512
509
455
Mean ISS, SD
456
Median pneumothorax size, IQR, mm
5.3 (8.6)
8.2 (16.5)
457
Significant (> 2 cm) hemothorax, %
12 (4.9)
5 (20)
< .01
458
Initial positive pressure ventilation, %
56 (22.2)
6 (24)
.84
513 514
459
Subsequent positive pressure ventilation, %
460
Presence of Respiratory distress, %
510
123 (49)
16 (64)
.15
124 (50.2)
14 (60.9)
.33
Presence of hemodynamic compromise, %
67 (27.2)
4 (18.2)
.36
463
Presence of decreased GCS, %
80 (32.3)
10 (40)
.43
464
Median hospital length of stay, IQR, d
10 (13.8)
11 (14.5)
.66
519
465
Median ICU length of stay, IQR, d
0 (4)
3 (12.0)
.15
520
466
Mortality, %
7.1
8
.88
461 462
467
515 516 517 518
521 522
ISS ¼ Injury Severity Score. See Table 1 legend for expansion of other abbreviations.
468
523
469
524
470
525
nonsignificant increased length of stay (11 vs 10 days; P ¼ .597). The two patients with mortalities in this group had an ISS score of 40 (> 25 represents severe or critical injuries) with intracranial hemorrhage, and it is unlikely that the pneumothorax contributed to the overall outcome. The rate of cardiothoracic surgery was
471 472 473 474 475 476 477 478 479 480
100.0
481 482
80.0
483
Frequency
484 485 486 488 489 490 491 492 493 494 495
print & web 4C=FPO
487
60.0 40.0 20.0 0.0 .00
10.00 20.00 30.00 40.00 50.00 60.00 Pneumothorax Size (mm)
Figure 2 – Pneumothorax size in conservatively managed patients.
higher in the intervention cohort, with 18 of 325 patients requiring surgery. Table 5 shows the characteristics of patients with a pneumothorax visible on chest radiography (overt pneumothorax) and of patients with a pneumothorax not visible on chest radiography (occult pneumothorax). One hundred seventy-seven patients had chest radiography as their initial chest imaging. One hundred thirty-seven of these patients proceeded to chest CT imaging. Of these 137 patients, 11 had a chest drain in situ at the time of chest radiography. Of the remaining 126 patients, 61 of 126 (48%) had no visible pneumothorax. Occult pneumothoraces were generally smaller than the overt pneumothoraces, with a respective median size of 7.26 mm vs 25.07 mm (P < .001). The majority, 470 of 603 study patients (78%), had evidence of rib fractures, with 361 of 470 patients and 143 of 470 patients having more than three and five rib fractures, respectively. Four hundred twenty-seven of 470 of the fractures were unilateral, and 167 of 470 were reported as flail chest, with 12 of 470 reported as
5
chestjournal.org
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
551 552
TABLE 3
] Hazard Ratios for Failed Conservative Management
606 607
Hazard Ratio
P Value
Male sex
1.05
.92
0.45-2.2
555
Size of pneumothorax ($ 2 cm vs < 2 cm)
1.61
.08
0.94-2.76
556
Bilateral vs unilateral pneumothorax
1.34
.25
0.83-2.12
611
557
ISS score (very severe vs severe and moderate severe)
1.17
.69
0.54-2.58
612
558
Presence of rib fractures
1.15
.57
0.71-1.88
613 614
553 554
559 560 561
Variable
Hemothorax (> 2 cm) Received initial positive pressure ventilation
95% CI
4.08
< .01
1.53-10.88
1.1
.84
0.44-2.76
608 609 610
615 616
562
Received subsequent positive pressure ventilation
2.10
.08
0.91-4.87
563
Presence of respiratory distress
1.23
.33
0.810-1.87
564
Presence of hemodynamic compromise
0.78
.37
0.45-1.34
619
565
Presence of decreased GCS score
1.17
.45
0.78-1.74
620
566 567
617 618
621
See Table 1 and 2 legends for expansion of abbreviations.
622
568 569 570 571 572 573 574 575 576 577 578 579 580 581 582
623
bilateral flail chest. There was no statistically significant association between the presence of rib fractures and significant hemothorax (c2 ¼ 0.946; P ¼ .331). There was a 10% complication rate associated with chest tube insertion. Fifteen patients (4.4%) required their drains to be resited, four patients (1.2%) had their drains dislodged, five patients (1.5%) had intraparenchymal drains on CT imaging, two patients (0.6%) experienced empyema, and one patient (0.3%) had a guidewire left in the pleural cavity. Eight patients (2.4%) required a subsequent drain after initial removal due to reaccumulation of air or fluid.
583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598
Discussion Chest drain insertion is not without risk of complications, with documented complication rates ranging from 15% to 30%.12,14-16 Current guidelines recommend chest tube placement for traumatic pneumothorax, particularly in patients receiving PPV,7 with a caveat that asymptomatic nonventilated patients can be managed with observation or aspiration at the treating clinician’s discretion. Existing literature has examined whether occult pneumothoraces can be managed conservatively.6,12 Scoring systems to determine whether chest tube intervention is required for occult pneumothoraces are in their infancy and have
599 600 601 602 603 604 605
TABLE 4
] Multivariable Cox Regression Analysis for Failure of Conservative Management
Variable Hemothorax > 2 cm
Hazard Ratio
P Value
5.29
< .01
95% CI 1.78-15.79
not been prospectively validated.17 We sought to determine whether traumatic pneumothoraces can be treated conservatively and examined factors that safely identify patients who could avoid chest tube insertion. In this paper, we show that the majority of patients managed conservatively, including patients receiving PPV, did not require further invasive ventilation. Studies have focused on whether there is a role for conservative management for occult pneumothoraces that are not initially visible on chest radiography.6,12 The resultant positive findings, including in patients receiving PPV, have been incorporated into clinical guidelines.18 Although these studies have been useful in establishing management pathways for traumatic pneumothoraces, they do have limitations. It is difficult to translate their findings into clinical practice in which CT is becoming the first-line investigation, with the majority of patients (70%) in our study having had a CT scan as their initial imaging and > 90% undergoing CT imaging during the inpatient stay. Furthermore, the distinction between overt and occult can be misleading. TABLE 5
Variable Age, SD, y
52.04 (23.80)
Overt Pneumothoraces (n ¼ 65) 56.30 (20.45)
631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647
48 (78.7) 7.26 (12.39)
651 652 653 654
47 (72.3)
655 656
25.07 (37.46)
657 658 20.00 (9.50)
16.00 (11)
659 660
See Table 2 legend for expansion of abbreviations.
6 Original Research
629 630
650
Occult Pneumothoraces (n ¼ 61)
Median ISS, IQR
627 628
649
Characteristics
Median size of pneumothorax on CT, IQR, mm
625 626
648
] Overt and Occult Pneumothoraces
Male sex, %
624
[
-#- CHEST - 2017
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
]
661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715
Whether a pneumothorax is seen on a chest radiograph is not solely related to its size and can be influenced by other factors, for example, the use of supine chest radiography has decreased sensitivity,8 with occult pneumothoraces in our study reaching > 80 mm in size. With this is mind, we proposed to look at the outcomes for traumatic pneumothoraces as a whole. Nearly half of the patients (46%) included were managed conservatively, with the majority of them (252 of 277 [90%]) not requiring subsequent invasive treatment for their pneumothoraces. This included 56 of 62 patients (90%) who received immediate PPV. Multivariable analysis supported the finding that immediate or subsequent PPV did not confer an additional risk of failure of conservative management. This is consistent with the most current study on occult pneumothoraces, with Moore et al12 demonstrating a failure rate of 14% in patients receiving PPV managed with observation. Although this was higher than their 4.5% failure rate for those not receiving PPV, the latter was not identified as an independent predictor of failed management on multivariate analysis, and no patient experienced tension pneumothorax related to delayed tube insertion. Smaller earlier studies found conflicting results. Brasel et al11 and Enderson et al9 demonstrated a 22% and 53% failure rate, respectively, for patients with occult pneumothoraces who were receiving PPV and were managed with observation. Brasel et al’s11 paper concluded that observation was safe in these patients, whereas Enderson et al9 recommended tube thoracostomy for all patients requiring PPV. This has led to an ongoing debate regarding the management of occult pneumothorax in patients receiving PPV. The East Practice Management Guidelines (2011) on occult pneumothoraces18 recommend that occult pneumothoraces may be observed in stable patients regardless of PPV. The currently recruiting Management of Occult Pneumothoraces in Mechanically Ventilated Patients (OPTICC) trial (ClinicalTrials.gov: NCT00530725), which is randomizing occult pneumothoraces in patients receiving PPV between chest tube insertion and observation, should contribute to this evidence base. The size of the pneumothorax was not a predictor of failed observation in our study on univariate and multivariate analyses, with nonsignificant differences in the size of pneumothoraces between the successful and failed observed groups. Pneumothorax size had previously been thought to be a predictor of progression,19 with De Moya et al17 proposing a scoring
system using the size of the occult pneumothorax and its relationship to the hilum to guide management. However, this has not been successfully validated,12 with Moore et al12 demonstrating that pneumothorax size was not an independent predictor of failed observation. The presence of a hemothorax appears to be predictive of failure of conservative management in both overt and occult pneumothoraces.12 This is consistent with clinical practice. A significant hemothorax is an indication for chest tube insertion to evacuate blood from the pleural space and avoid complications such as infection and fibrothorax, and when combined with the presence of a pneumothorax, this provides a strong incentive for intervention. In this study, patients who did not require prehospital or admission chest procedures generally did not require a chest procedure later during hospitalization. When this information is combined with previous trials on traumatic pneumothoraces, it appears that there is a subpopulation that can be managed conservatively. Certainly, when there is no significant hemopneumothorax (< 2 cm), there can be consideration for expectant management. Mechanism of injury, ISS, or size of the pneumothorax does not appear to provide a strong indication for intervention. Additionally, in our study, clinical condition did not confer an adverse prognosis, although in other studies, respiratory distress did.12 Although the use of ventilation has been controversial, it appears from our findings and previous studies that pneumothoraces can be managed conservatively with careful observation in patients receiving PPV with no increased risk of harm.6,12
716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752
This is a retrospective observational trial and as such is subject to the inherent limitations of such a study. The data were collected from a single center with a low rate of penetrating chest wall injuries (5%), which should be considered when generalizing the findings to other centers. Selection bias may have been introduced by physician selection and variation in initial imaging modality, and the decision to intervene may have affected the conservatively treated cohort characteristics, and it is likely that high-risk unwell patients were underrepresented in the conservatively treated arm. Those treated with an immediate intervention, despite a similar ISS, likely represented a more unwell population, with higher rates of cardiorespiratory compromise, PPV use, surgical referral rates, and mortality. The length of stay criteria (length of stay ¼ 3 days or more, or admittance to a high-dependency area regardless of
7
chestjournal.org
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770
771
length of stay) is likely to have been biased against patients conservatively managed successfully and not requiring a prolonged hospital admission, suggesting that the overall rate of effective conservative management is probably greater. Efforts were made to minimize bias by including a large number of consecutive unselected patients into the analysis and careful documentation and comparison of cohort characteristics.
772 773 774 775 776 777 778 779 780 781 782
Conclusions
783
Our study represents the largest observational study on traumatic pneumothoraces to date. It demonstrates that
784
the majority of conservatively managed patients were successfully managed without requiring a chest drain. This includes the majority of patients receiving PPV, the use of which did not present an increased risk of failure of expectant management. This study provides support for an observational expectant approach if the treating physician does not believe that an immediate chest drain is warranted in a patient with a traumatic pneumothorax. Future prospective randomized trials examining the outcomes of a conservative approach in traumatic pneumothorax, regardless of pneumothorax size or use of PPV, would help clarify which patients are best managed expectantly.
785
826 827 828 829 830 831 832 833 834 835 836 837 838 839 840
786
841
787
842
788
Acknowledgments
789
Author contributions: S. P. W. takes responsibility for the content of the manuscript, including the data and analysis. S. P. W. performed statistical analysis and prepared the manuscript. S. P. W., S. L. B., J. T., and N. A. M. conceived the design of the study. All authors read and approved the final manuscript for submission.
790 791
Q14
792 793 794 795
Q12
796 797 798 799
References
800
1. Di Bartolomeo S, Sanson G, Nardi G, Scian F, Michelutto V, Lattuada L. A population-based study on pneumothorax in severely traumatized patients. J Trauma Acute Care Surg. 2001;51(4):677-682.
801 802 803 804 805
2. Idris BM, Hefny AF. Large pneumothorax in blunt chest trauma: is a chest drain always necessary in stable patients? A case report. Int J Surg Case Rep. 2016;24:88-90.
806 807 808
3. Kshettry VR, Bolman R, 3rd. Chest trauma. Assessment, diagnosis, and management. Clin Chest Med. 1994;15(1):137.
809 810 811 812 813 814
Financial/nonfinancial disclosures S. L. B. has received funding for accommodation and registration fees from Boehringer Ingelheim. None declared (S. P. W., J. T., N. M.).
Q13
4. Centers for Disease Control and Prevention NCfHSN. National hospital discharge survey: 2007 summary. National health statistics reports, no. 29. National health statistics reports, no. 29. Atlanta, GA: National Center for Health Statistics; 2010.
5. Centers for Disease Control and Prevention. Web-based Injury Statistics Query and Reporting System (WISQARS), 2007. http://www.cdc.gov/ injury/wisqars. Accessed xxx. 6. Wilson H, Ellsmere J, Tallon J, Kirkpatrick A. Occult pneumothorax in the blunt trauma patient: tube thoracostomy or observation? Injury. 2009;40(9):928-931. 7. ATLS Subcommittee; American College of Surgeons’ Committee on Trauma; International ATLS working group. Advanced trauma life support (ATLSÒ): the ninth edition. J Trauma Acute Care Surg. 2013;74(5):1363. 8. Ball CG, Kirkpatrick AW, Feliciano DV. The occult pneumothorax: what have we learned? Can J Surg. 2009;52(5):E173. 9. Enderson BL, Abdalla R, Frame SB, Casey MT, Gould H, Maull KI. Tube thoracostomy for occult pneumothorax: a prospective randomized study of its use. J Trauma Acute Care Surg. 1993;35(5): 726-730. 10. Plurad D, Green D, Demetriades D, Rhee P. The increasing use of chest computed tomography for trauma: is it being overutilized? J Trauma Acute Care Surg. 2007;62(3):631-635. 11. Brasel KJ, Stafford RE, Weigelt JA, Tenquist JE, Borgstrom DC. Treatment of occult pneumothoraces from blunt trauma. J Trauma Acute Care Surg. 1999;46(6):987-991.
12. Moore FO, Goslar PW, Coimbra R, et al. Blunt traumatic occult pneumothorax: is observation safe?—results of a prospective, AAST multicenter study. J Trauma Acute Care Surg. 2011;70(5): 1019-1025. 13. Ryan MT, Caputo ND, Lakdawala V, Jara F. Spontaneous resolution of a large traumatic pneumothorax. Am J Emerg Med. 2012;30(5):833.e3. 14. Etoch SW, Bar-Natan MF, Miller FB, Richardson JD. Tube thoracostomy: factors related to complications. Arch Surg. 1995;130(5):521-526.
843 844 845 846 847 848 849 850 851 852 853
15. Bailey R. Complications of tube thoracostomy in trauma. J Accid Emerg Med. 2000;17(2):111-114.
854
16. Ball CG, Lord J, Laupland KB, et al. Chest tube complications: how well are we training our residents? Can J Surg. 2007;50(6):450.
856
855
17. De Moya MA, Seaver C, Spaniolas K, et al. Occult pneumothorax in trauma patients: development of an objective scoring system. J Trauma Acute Care Surg. 2007;63(1):13-17. 18. Mowery NT, Gunter OL, Collier BR, et al. Practice management guidelines for management of hemothorax and occult pneumothorax. J Trauma Acute Care Surg. 2011;70(2):510-518. 19. Garramone R, Jr, Jacobs L, Sahdev P. An objective method to measure and manage occult pneumothorax. Surg Gynecol Obstet. 1991;173(4):257-261.
815
857 858 859 860 861 862 863 864 865 866 867 868 869 870
816
871
817
872
818
873
819
874
820
875
821
876
822
877
823
878
824
879
825
880
8 Original Research
[
-#- CHEST - 2017
FLA 5.5.0 DTD CHEST1397_proof 15 November 2017 2:25 am EO: CHEST-17-0836
]