A Study of Cases in Which Penrose Drains Were Selected for Postoperative Hemorrhaging After Head and Neck Surgery

Abstract

Closed suction drains are commonly used after head and neck surgery, but airway obstruction due to postoperative hemorrhaging can be fatal. Some studies have suggested that cervical circumference should be assessed after thyroid surgery to detect such problems early. Rapid responses may be difficult at facilities with limited staffing. We opted for Penrose drains in cases in which the risk of postoperative hemorrhaging was expected to be high, and we examined the effectiveness of this approach in this study. Between October 2022 and December 2023, 25 of 68 (37%) head and neck operations involved the placement of a Penrose drain. Their medical records were reviewed retrospectively. There were 20 cases involving neck dissection, which was performed via a wide range of surgical techniques. Intraoperative findings (e.g., significant oozing of blood) was the most common reason for choosing the open technique (10 cases). The duration of drain placement was < 7 days (mean: 4.68 days). There were no surgical site infections; two cases of postoperative hemorrhaging; and no fatal problems, such as airway obstruction, but delayed wound healing due to wound-edge dehiscence was observed in some cases. There were no cases of airway narrowing due to postoperative hemorrhaging. Penrose drains are considered a good indication for cases with a history of radiotherapy causing postoperative haemorrhage, cases on oral anticoagulants, lesions in the deep parotid to parapharyngeal space that are difficult to recognise for haematoma formation, and cases where negative pressure should be avoided for exposure of anastomotic vessels and nerves, and should continue to be investigated.

Introduction

Closed suction drains are commonly used after head and neck surgery, but airway obstruction due to postoperative hemorrhaging can be fatal. We have had cases in which we have placed a total of eight closed suction drains, four per side, e.g., in cases of bilateral neck dissection. However, even when using such a high number of drains hematoma formation and open wounds due to postoperative bleeding could not be completely prevented. Some studies have suggested that the cervical circumference should be assessed to facilitate the early detection of postoperative hemorrhaging after thyroid surgery [1]. Furthermore, it can be difficult for clinicians at less well-staffed facilities to respond quickly to postoperative hemorrhaging. Therefore, for several years we have opted for Penrose drains in cases where the risk of postoperative hemorrhaging was expected to be high. In this study, we reviewed the effectiveness of this approach and any associated problems.

Subjects and Methods

Between October 2022 and December 2023, Penrose drains were placed during 25 of 68 head and neck operations (37%). The patients were 24 males and one female, who were aged between 48 and 88 years (median: 71 years). The primary site, procedure, reasons for choosing the Penrose drain, duration of drain placement, and adverse events were reviewed retrospectively by accessing the patients’ medical records. Length of hospital stay was not considered, as some patients required rehabilitation for post-operative dysphagia until oral intake was possible, and some required post-operative chemo-radiotherapy. The study was approved by the ethics committee of Miyazaki Prefectural Miyazaki Hospital (approval no. 24-43).

Results

The primary site was the oropharynx in seven cases; the oral cavity in five cases; the hypopharynx or larynx in three cases each; the thyroid gland or parotid gland in two cases each; and the external auditory canal, maxillary sinus, or nasal cavity in one case each (Table 1). There were 20 cases involving neck dissection, which was performed via a wide range of surgical techniques. Intraoperative findings (e.g., a significant amount of oozing blood) (Fig. 1) was the most common reason for choosing a Penrose drain (10 cases), followed by a history of radiotherapy (6 cases); antiplatelet medication (3 cases); blood loss requiring transfusion, re-operations, and complications, such as heart failure, in two cases each; and parapharyngeal space lesions in one case (Fig. 2). The drains were kept in place for < 7 days (mean: 4.68 days). There were no cases in which the objective was not achieved due to the drain dropping out during the procedure. There were also no cases of surgical site infections. There were two cases of postoperative hemorrhaging, both involving patients with a history of radiotherapy. No fatal problems, such as airway obstruction, occurred, but delayed wound healing due to wound-margin dehiscence was observed (Fig. 3).

Table 1.

Summary of clinical features in Penrose drain inserted subjects

Subject ID	Age	Gender	Primary site	Surgical method	Reason for open drainage system	Drainage period * (mean POD 4.68)	Postoperative bleeding	SSI
1	58	M	Oral	Pull through + ND + FA	Impression during surgery (oozing, etc.)	4	No	No
2	77	M	Oral	Pull through + ND + FA	History ofRT or CRT	6	No	No
3	65	M	Oral	Pull through + ND + FA	Transfusion (bleeding > 600 ml)	7	No	No
4	62	M	Oral	pull through + ND + PMMC	history ofRT or CRT	5	No	No
5	75	M	Oropharynx	Pull through + ND + PMMC	Impression during surgery (oozing, etc.)	7	No	No
6	76	M	Oropharynx	Pull through + ND + PMMC	History ofRT or CRT	5	No	No
7	70	M	Oropharynx	Pull through + ND + RAMC	Impression during surgery (oozing, etc.)	4	No	No
8	72	M	Oral	ND	Complications (heart failure, diabetes mellitus, etc.)	5	No	No
9	65	M	Nasal cavity	ND	Impression during surgery (oozing, etc.)	5	No	No
10	78	M	Hypopharynx	ND	Antiplatelete drugs	7	Yes**	No
11	82	M	Thyroid	ND	Post surgical history	2	No	No
12	71	M	Oropharynx	TOVS + ND	Impression during surgery (oozing, etc.)	6	No	No
13	88	M	Oropharynx	TORS + ND	Impression during surgery (oozing, etc.)	3	No	No
14	50	M	Oropharynx	TORS + ND	Impression during surgery (oozing, etc.)	2	No	No
15	48	M	Oropharynx	TORS + ND	Impression during surgery (oozing, etc.)	2	No	No
16	58	M	Larynx	TL + ND	Impression during surgery (oozing, etc.)	3	No	No
17	68	M	Larynx	TL + ND	Complications (heart failure, diabetes mellitus, etc.)	4	No	No
18	83	M	Hypopharynx	TPL + ND + FJF	Post surgical history	7	No	No
19	65	M	Hypopharynx	TPL + ND + FJF external tempral bone	transfusion (bleeding > 600 ml)	7	No	No
20	74	F	External auditory canal	resection + superficial lobe parotidectomy + ND	Impression during surgery (oozing, etc.)	6	No	No
21	87	M	Parotid gland	Total parotidectomy + PMMC	antiplatelete Drugs	7	No	No
22	54	M	Parotid gland	Deep lobe parotidectomy	Deep lesion (deep lobe parotidectomy)	5	No	No
23	71	M	Maxilla	Maxillectomy + RAMC	History of RT or CRT, antiplatelete drugs	4	No	No
24	76	M	Larynx	Vartical partial laryngectomy	History ofRT or CRT	2	No	No
25	77	M	Thyroid	Hemi-thyroidectomy	History ofRT or CRT	2	Yes**	No

POD: Post Operative Day, SSI: Surgical Site Infection, ND: neck dissection, FA: free forearm flap, PMMC: pectoralis major myocutaneous flap, RAMC: rectus abdominis　musculocutaneous flap, TOVS: transoral videolaryngoscopic surgery, TORS: transoral robotic surgery, TL: total laryngectomy, TPL: total laryngopharyngectomy, FJF: free jejunal flap, RT: radiation therapy, CRT: chemoradiation therapy

Fig. 1.

A case involving lateral temporal bone resection, superficial parotidectomy, and supraomohyoid neck dissection. A Intraoperative findings. B Findings at the time of closure. We chose a Penrose drain in this case because blood was oozing from the osteotomy edge (arrow) and the facial nerve (arrow head) was exposed

Fig. 2.

Case of a lesion in the region from the parotid deep lobe to the parapharyngeal space (case 22). A MRI T2WI coronal section. Mass in the deep lobe of the left parotid gland to the parapharyngeal space. B Intraoperative findings. Tumour on the dorsal surface of the facial nerve (arrow). C Intraoperative findings. Facial nerve in the vicinity of the Penrose drain (arrow). D Findings at the time of closure. When deciding which type of drain to use, we considered the risk of pharyngeal edema resulting from the delayed detection of a hemorrhage due to the deep location of the lesion, leading to airway obstruction

Fig. 3.

In two patients with a history of chemoradiotherapy resulting in postoperative hemorrhaging, wound-margin dehiscence, and delayed wound healing due to inadequate fusion occurred after hematoma removal and hemostatic procedures. A A case involving right lobectomy of the thyroid gland (Case 25); B A case of right neck dissection (Case 10)

The two cases of postoperative hemorrhaging are presented below. Case 25: This case involved right lobe thyroidectomy being performed in a patient that had previously received chemoradiotherapy to the neck. The intraoperative oozing of blood from the scar tissue was noticeable, and the wound was closed with a Penrose drain, which was placed in case postoperative hemorrhaging occurred. The operation lasted 2 h 18 min, and there was 39 g of intraoperative blood loss. The patient returned to the room at 16:40 h. Gauze contamination of the wound with blood was minimal that day. However, gauze contamination with fresh blood became noticeable at 1:00 the following morning, and the bleeding volume at this point was 30 g. At 2:00, 50 g of gauze hemorrhaging was recorded. At 4:40, 50 g of gauze hemorrhaging was reported to the on-call doctor. At 06:00, the on-call doctor (resident) arrived and noted 80 g of gauze hemorrhaging. A total of 300 g of blood was lost in the six hours leading up to the arrival of the surgeon at 07:00. Hemostasis was achieved by opening the wound and removing a clot. The source of the bleeding was unknown, and blood was thought to be oozing out of a wound due to an increase in the levels of fibrinolytic system components around the hematoma, rather than due to an arterial injury. There was no cervical swelling during this period and no change in the patient’s vital signs. The hemoglobin level decreased from 12.9 g/dl preoperatively to 9.7 g/dl. After haemostatic treatment, the wound was filled with ointment gauze as an open wound, and the gauze was removed the following day to ensure that there was no bleeding and sutured closed. The patient's general condition was good, but he was discharged on post-operative day 9 due to family pick-up availability. During an outpatient consultation on post-operative day 29, a wound margin dehiscence was identified and a policy of moist therapy with ointment application and gauze covering was adopted to wait for wound margin healing and epithelialisation. During an outpatient visit on post-operative day 49, wound margin fusion and epithelialisation were confirmed. Case 10: This was a case of neck dissection for late right cervical metastasis, involving a patient who had previously received radiotherapy for hypopharyngeal cancer. The patient had a history of taking direct-acting oral anticoagulants (DOACs) due to ischemic heart disease, which were discontinued preoperatively. The wound was closed with a Penrose drain, which was placed in case of postoperative hemorrhaging. The operation lasted 2 h 28 min, and there was 68 g of intraoperative blood loss. The patient returned to the ward from the operating theatre at 17:30. The following morning, the patient complained that he had experienced difficulty in gaining strength in his limbs since the operation, and an MRI scan revealed a hyperacute cerebral infarction extending from the medial cortex to the medulla of the left frontal lobe. A neurologist was consulted, the DOAC treatment was resumed, and treatment with a combination of two antiplatelet drugs was started. A small amount of gauze blood contamination appeared on postoperative day 4, but the drain was removed on postoperative day 7 due to a small amount of persistent gauze bleeding. A small amount of oozing from the drain extraction site persisted on postoperative day 8. On postoperative day 9, the patient required gauze changes every 1–2 h. On postoperative day 10, the wound was opened, and hematoma removal was performed to stop bleeding. The patient’s hemoglobin level at this time was 14.4 g/dl, which was not significantly different from their preoperative level of 14.5 g/dl. No changes in vital signs or airway emergencies occurred during this period. Suturing of the wound margin after haemostatic treatment was difficult due to skin hardening and shrinkage, and a policy of moist therapy with ointment application and gauze covering was adopted to wait for healing and epithelialisation of the wound margin. The patient was discharged on post-operative day 19, as it took several days to teach and learn self-treatment due to the patient's advanced age. Outpatient Trafermin spraying and ointment treatment was continued, but it took 10 months for the wound to epithelialise.

Discussion

The present study is not intended to demonstrate the superiority or non-inferiority of Penrose drains over closed suction drains. We would like to suggest that the use of Penrose drains should be actively considered in selected cases, rather than closed suction drains being used uniformly.

Hemorrhaging is the most important adverse event to avoid after head and neck surgery, as airway obstruction caused by a hematoma can lead to serious sequelae, such as hypoxic encephalopathy, or even be fatal [2, 3]. The drains used to avoid this are generally of the closed-suction type. We have had cases, for example, of bilateral neck dissection, in which four closed suction drains were placed on each side, resulting in a total of eight closed suction drains. However, even when this number of drains was used, hematoma formation and open wounds due to postoperative bleeding could not be completely prevented. The risk associated with hematoma formation; i.e., that bleeding that exceeds the negative pressure of a drain may lead to airway narrowing, means that wound drainage cannot be avoided. During such open surgical procedures, we have chosen to use Penrose drains instead of closed suction drains. We considered that with this approach even if heavy bleeding occurred, it was not likely that blood would accumulate in the tissue and produce a hematoma. The subsequent course of events was uneventful in all cases. Therefore, we considered that Penrose drains should be chosen from the outset in cases where the risk of postoperative hemorrhaging appears to be high. In the present study, there were two cases of postoperative hemorrhaging in cases of Penrose drain placement, but no airway narrowing due to hematoma formation was observed, despite a total of more than 300 g of blood loss occurring in one case, which required six hours to treat.

Some studies have advocated cervical perimetry [1] in the postoperative period after thyroid surgery in order to ensure the early detection of postoperative hemorrhaging. However, it is difficult to apply this approach in cases involving deep lesions, such as lesions in the region from the parotid deep lobe to the parapharyngeal space, and there is concern that the detection of hemorrhages in this region may be delayed. Indeed, some studies have suggested that deep cervical bleeding may not result in cervical swelling [4]. Therefore, a Penrose drain is a good option in such cases.

In thyroid surgery, there is debate about the need for drains in the first place. In cases of goiter without mediastinal spread, involving patients with normal coagulation indices that undergo surgery that does not require lateral neck dissection, there is no clear evidence that the use of drains significantly improves patient outcomes, and some systematic reviews have concluded that drains may be associated with longer hospital stays [5]. However, in some cases, such as those we experienced, if 300 g of blood were to accumulate under the anterior cervical skin, airway narrowing could occur, and the placement of a Penrose drain did not result in serious consequences in any of our cases. Therefore, efforts should be made to select the best response (no drain, closed suction drain, or Penrose drain) in individual cases without blindly following the evidence in the literature.

The main reason for the common use of closed suction drains is to prevent surgical site infections, as recommended in the CDC guidelines [6]. In other types of surgery, comparisons between closed suction drains and Penrose drains have shown the former to be significantly superior in terms of postoperative infections [7, 8]. However, there were no cases of infection due to the use of Penrose drains in the present study. Note that we used antimicrobials according to our clinical pathway for oral, pharyngeal, and laryngeal head and neck surgery, with ABPC/SBT administered preoperatively and intraoperatively, as well as until postoperative day 5. Another major reason for the use of closed suction drains may be the influence of previous studies on wound healing. A study examining whether there is a difference in neck healing after neck dissection between active and passive drains reported that patients treated with passive drains had a delayed healing rate of 54%, whereas the delayed healing rate of the patients treated with active drains was 6%, and stated that active drains should be used after neck dissection [9]. However, in our study no cases of delayed wound healing occurred, except in two cases in which hematomas were treated.

The main advantage of closed suction drains is that there is less risk of obstruction due to misalignment, dislodgement, or flexion compared with Penrose drains [10]. However, in our study there were no cases of drains dropping out or problems due to blockages. Other advantages of closed suction drains include the fact that they allow easy observation of the wound and draining fluid, eliminate the need for frequent wound care, and reduce the burdens on patients and healthcare professionals. It is true that Penrose drains may require frequent changes of contaminated gauze, but it is possible to assess the situation by recording the amount (weight) and nature of drainage fluid and obtaining photographs at each examination, as was done in our bleeding cases. Some studies have cited the accelerated wound-healing effect of negative-pressure loading [11] as an advantage of closed suction drains. We have also had the opportunity to experience vacuum-assisted closure therapy [12] for fistulas and refractory ulcers due to suture failure. However, in the present study drain removal and wound margin-healing were achieved within a week in all cases except the two that required treatment for postoperative hemorrhaging; therefore, it seems unlikely that significant differences in postoperative drainage would occur in the short term.

On the other hand, the disadvantages of closed suction drains include the difficulty of coping with bleeding that exceeds the negative pressure of the drain, as well as the non-negligible risk of damage to nerves and anastomotic vessels in the wound due to suctioning. In this respect, cases involving parotid (facial nerve) or thyroid (recurrent nerve) surgery and exposed nerves, or free flaps with microvascular anastomoses are considered good indications for Penrose drains (Fig. 1). Indeed, there have been reports of cases in which anastomotic vessels have been sucked into closed suction drains [13].

The only disadvantage of Penrose drains we encountered was that in cases where the skin had hardened, such as after chemoradiotherapy, there was little room for extension at the wound margins, and loose suturing after hematoma removal may lead to a risk of wound-edge break-up and delayed wound healing (Fig. 3). Wound healing took 1.5 months in case 26 and 10 months in case 10, although this period could have been reduced with VAC therapy, etc. Self-treatment with ointment applied to the wound allowed outpatient management without significantly compromising quality of life.

One limitation of this study is that it was a single-center study involving a small number of cases. Another is that the two cases of postoperative hemorrhaging experienced in this study both involved relatively weak bleeding; i.e., the bleeding was considered to be of venous origin with no clear bleeding point. In these cases, the bleeding appeared to originate from around a small pool of coagulum in the wound, which gradually decreased after the removal of the coagulum, and the bleeding stopped without the use of ligation or electrocoagulation. It has long been shown in studies of chronic subdural hematomas that the fibrinolytic system is hyperactive around hematomas, resulting in exacerbation of the hematoma [14]. In our cases, the bleeding is assumed to have stopped due to the removal of the cause. There have been some reports of fatal hemorrhaging after thyroid surgery [15], but all of these cases were due to the disruption of large blood vessels, such as the common carotid artery or brachiocephalic artery, after the second postoperative day, which was considered to be caused by infection (except in one case when the hemorrhaging occurred 1.5 h postoperatively due to a ligature around the superior thyroid artery becoming dislodged). However, we have not experienced any such cases. Further experience is needed to determine whether such hemorrhaging can be managed with a Penrose drain without causing an airway emergency. However, initially such massive hemorrhages should be considered to involve the deterioration of circulatory dynamics rather than represent an airway emergency due to the formation of a hematoma, and they are not considered to be affected by drainage properties. In our experience, arterial postoperative hemorrhaging is most often noticed immediately after surgery, especially if neck swelling occurs before and after extubation or there is a sudden increase in the volume of a closed negative-pressure drain bag, in which case the wound should be opened immediately, and the arterial bleeding point should be identified and ligated. However, in a case in which bleeding occurred several hours after a patient returned from the operation room, diffuse venous hemorrhaging was observed from within the wound, but hemostasis was achieved by removing a clot, and no pulsatile bleeding points were identified. Therefore, we consider that bleeding that occurs after the immediate postoperative period can be managed with a Penrose drain, unless a major blood vessel has been ruptured.

Another limitation of this study is that it is difficult to establish clear selection criteria for Penrose drains from the sample size of the present study. Therefore, based on our experience with the present results and the aforementioned literature review, we devised a tentative selection criteria for Penrose drains (Table 2). It should be considered an active criterion of choice in cases of previous radiotherapy and oral antiplatelet and anticoagulant medication, where we have experienced postoperative haemorrhage, as well as in cases of deep neck lesions such as the parapharyngeal space, where detection can be delayed if it occurs, and where there is a risk of accidental aspiration of critical structures such as nerves and anastomotic vessels with closed suction drains. We intend to continue our review accordingly.

Table 2.

Tentative selection criteria for Penrose drains

Applicable conditions	Reasons
History of radiotherapy	Hemorrhagic
History of taking anticoagulants or antiplatelet drugs	Hemorrhagic
Parapharyngeal space lesions	Hematoma formation is difficult to notice
Nerve exposures and anastomotic vessels	Negative pressure should be avoided

Conclusion

Twenty-five cases in which Penrose drains were selected for postoperative hemorrhaging after head and neck tumor surgery were studied. There were no cases of airway narrowing due to postoperative hemorrhaging. Penrose drains are considered a good indication for cases with a history of radiotherapy causing postoperative haemorrhage, cases on oral anticoagulants, lesions in the deep parotid to parapharyngeal space that are difficult to recognise for haematoma formation, and cases where negative pressure should be avoided for exposure of anastomotic vessels and nerves. Instead of uniformly using closed suction drains, it is advisable to consider using Penrose drains in selected cases.

Declarations

Conflict of interest

The authors have no conflicts of interest to declare.

Informed consent

This was a retrospective study; therefore, formal consent was not required.