To our knowledge, this is the largest study assessing a single surgeons practice with timing of parotid drain removal, and the first study to survey Canadian otolaryngologists about their practices surrounding parotid drain management.

There is no current clear consensus or North American guideline on parotid drain management. Outpatient parotidectomies have been described since 1991 [5], but provider variability in post-operative management remains and some patients stay in hospital until the drain is removed. This differs from other procedures in otolaryngology, such as thyroidectomies, where same-day discharges are recommended by governing bodies as safe in the correct patient population [6]. Factors that have been previously attributed to the success of outpatient thyroidectomies in otolaryngology include lack of significant patient co-morbidities, patient proximity to hospital, and availability of social support [7].

Based on our ten-year retrospective review, the only significant demographic difference between the patients who had their drains removed four hours post-operatively compared to the patients who had their drains removed POD ≥ 1 was their age (50 vs. 56 years respectively, p < 0.001) (Table 1). This may be due to older patients generally being more comorbid, which may require admission to hospital secondary to surgical, anesthetic, or social concerns. If patients were already scheduled to be admitted after the practice change to POD 0 drain removal, the patient’s drain would be left for removal on POD ≥ 1 as they would still be in hospital regardless. Other studies have also found patients who were kept in-hospital after parotidectomy were significantly older [8]. Chen et al.’s study on post-operative drain output after parotidectomies found that body weight was the only patient demographic factor significantly associated with increased post operative drain output [9].

In our 526 patients, the patients who had their drains removed POD ≥ 1 did have longer lengths of stay, as expected. There was no difference in development of a hematoma or seroma in early drain removal at four hours post-operatively (POD 0) compared to drain removal POD ≥ 1. Although there was a trend towards a higher hematoma formation rate in the POD 0 cohort than the POD ≥ 1 cohort, the limited power of this 10-year retrospective study could not bring out any statistically significant difference in hematoma complication rates (p = 0.15).We also did not find any statistically significant difference in seroma rates between the two cohorts (Table 3).

Roh and colleagues demonstrated that parotid gland salivary flow rate was higher after partial parotidectomy compared to conventional parotidectomies [10]. The amount of saliva and potential for sialocele formation are considerations providers must consider when debating drain management after parotidectomies [11]. In our study, there was no difference in hematomas or seromas when assessing superficial vs. deep/complete parotidectomies. Our hematoma and seroma rates for patients with same-day drain removal (2.6% and 2.1% respectively) are similar to other reported values in the literature of 3.8–6.1% and 2–10% [3, 12,13,14].

Being aware of variables such as anticoagulation use, coagulopathy, and uncontrolled hypertension (Table 3) that may play a role in parotid hematoma development could help providers choose which patients need longer drain management. We found no significant difference in hematoma or seromas based off malignant versus benign pathology. Molfe and Urquhart, in contrast, examined 69 patients who underwent superficial parotidectomies and found malignant pathology was significantly associated with increased post-operative drain output [3].

An analysis of the American College of Surgeons National Surgical Quality Improvement Program reported an overall total complication rate of 5.3% after parotidectomy [15]. Our infection rate of 9.7% was higher than the rates reported of 3.8–5.4% [12, 16] but there was no significant difference when comparing the infection rates between the POD 0 and POD ≥ 1 cohort. We defined infection as patients who received antibiotics, and this broad definition as well as the lack of perioperative prophylactic antibiotics may contribute to our higher infection rate.

Given the low incidence of hematoma formation (2.6%) and the shortened hospital stay in patients who have drains removed at four hours (0.40 days vs. 1.08 days, p < 0.001), the senior surgeon has chosen to continue the practice of removing drains prior to same day discharge.

Overall, our survey response rate was approximately 21.5%, with 176 responses. There is high variability in online surveys of surgeons, with reports ranging from 9 to 80% [17]. Our response rate is comparable to recent Canadian national surveys in Otolaryngology that range from 22 to 30% [18,19,20]. To our knowledge, this is the first North American examination of practice in drain management after parotidectomy.

Of our respondents, just under half (47.8%) performed between 10 and 30 parotidectomies annually. The majority of responses (62.8%) favored closed drains, which is also the preference of our senior surgeon. Of those using a drain output based criteria for drain removal, most stated their criteria was ≤ 30 ccs in 24 h or ≤ 20 ccs in 8 h (Table 4). For those that favored a time based criteria for drain removal, the most common response was removal on POD1 (66.1%) followed by POD2 (12.5%). Comparing these values to other reports in the literature, there are studies that both report higher and lower drain output cut-offs for removal. Harris and colleagues assessed the timing of drain removals after parotidectomies and found a safe profile in removal when the volume was less than or equal to 50 mL after 24 h [21]. In comparison, another study removed drains when the output was less than 5 mL after 8 h [3].

There were three participants who stated they never used drains and twelve who stated they ‘rarely used drains (< 10%)’. A large recent study from Denmark of 205 patients undergoing superficial parotidectomy compared those with drain outputs less than 25 mL compared to those with more than 25 mL in 24 h [22]. The authors reported that 7.3% of patients developed seromas or hematomas in spite of drain placement and that the choice of placing a drain was not significantly associated with drain output.

Medicine has moved towards less invasive procedures and minimizing hospital stays to improve patient outcomes and curb accelerating healthcare costs. A retrospective review of 42 drainless patients compared to 49 patients with drains after parotidectomy supports the argument for early discharges. The authors found a trend towards more seromas in the patients without drains, but this was not significant (p = 0.298) [23]. None of their drainless patients required readmission or experienced major complications. Our study, in comparison, also assessed the practice of minimizing healthcare resources by reducing admissions and routinely removing drains POD 0 with same day discharge. The seroma and hematoma rate in our patients included also trended higher in the POD 0 group than the POD ≥ 1 group, but this was also not significant.

Recent systematic reviews also support the safety of outpatient parotidectomies, with similar complication and re-admission rates when compared to inpatient parotidectomies [24, 25]. Lee et al. retrospectively examined 238 patients who underwent superficial parotidectomy and found no difference in complication rates, return to the emergency department, or readmission within 30 days after outpatient vs. inpatient post-operative stays [8]. Their cut-off for drain removal was < 30 mL in 24 h. Similar to our study, the found that the inpatient cohort who stayed overnight were statistically older than the outpatient cohort.

Limitations of this study include the inherent bias associated with a retrospective review and the selection bias of providers who choose to answer the survey on drain management. Our cohort included heterogenous pathologies (benign and malignant cases), although there were no differences in hematoma or seromas based off pathology. Unfortunately, we were also not able to assess the exact amount of the parotid tissue removed, which may have made a difference when assessing drain usage or post-operative complication rates. Furthermore, the authors do acknowledge that this study examines a single surgeons practice and that the results may not be generalizable. Some practitioners discharge their patients with drains in place for removal at in-office follow-up, which would still decrease hospital stays. However, these providers then do require follow-up, which may be inconvenient to patients. Lastly, the national survey data focuses on providers current practice but did not specifically ask the reasoning of providers, which would have provided additional information. Our response rate for the survey data was low (21%), but comparable to other recent national surveys of Canadian Otolaryngologists (22–30%) [18,19,20].