Predictors of hypoxemia after general anesthesia in the early postoperative period in a hospital in Ethiopia: an observational study.

BACKGROUND: Early postoperative hypoxemia is a common problem after general anesthesia. The identification of factors associated with an increased occurrence of it might help healthcare professionals to hypoxemia risk patients, therefore this study aims to assess the incidence and factors associated with early postoperative hypoxemia among surgical procedures.
METHODS: A prospective cohort study design was conducted from February 1, 2020 to June 30, 2020, on a total of 424 patients who underwent surgery under general anesthesia in Debre Tabor Comprehensive Specialized Hospital. The data was collected using a structured checklist. Bivariable and multivariable logistic regressions were used to check the association.
RESULTS: The incidence of early postoperative hypoxemia was 45.8%. Patients having a BMI of 25-29.9 kg/m2 and BMI of 30-39.9 kg/m2, patients having a chronic disease, current smokers, SPO2 reading before induction of less than 95%, emergency surgery, and the absence of oxygen therapy during the period of transfer and/or in the post anesthesia care unit were significantly associated with an increased risk of hypoxemia in the early postoperative period.
CONCLUSIONS: The incidence of early postoperative hypoxemia was high in Debre Tabor Comprehensive Specialized Hospital. Obese patients, patients having a chronic disease, current smokers, and lower oxygen saturations before induction, emergency surgery, and the absence of oxygen therapy were the main predictors of an increased occurrence of early postoperative hypoxemia. ©Copyright: the Author(s).

Introduction

The early postoperative period is considered to be the critical period for the development of hypoxemia [1,2]. This occurrence is common in the post-anesthesia care unit after major abdominal, orthopedic, gynecologic, and obstetrics procedures [1]. It is not restricted to the seriously ill patients but may occur in otherwise healthy ones [3,4]. Postoperative hypoxemia occurred secondary to impaired gas exchange during anesthesia as a result of reduced tone in the muscles of the chest wall persisting into the postoperative period [5,6]. Pulse oximetry is commonly used monitoring to diagnose hypoxemia. Normal pulse oximeter readings should always be 95 to 100% during anesthesia and recovery from anesthesia. If the oxygen saturation is 94% or lower, the patient is hypoxemic and needs to be treated quickly. A saturation of less than 90% is a clinical emergency in postoperative patients. Diagnosis of hypoxemia with the help of a pulse oximeter would lead to early correction of the hypoxemic event [7].

Despite recent advances in anesthesia and surgical care, the incidence of hypoxemia has remained high because of the increasing complexity of the surgery and condition of patients [8,9]. The incidence of hypoxemia is highest during the early postoperative period which results in worse outcomes such as myocardial ischemia, organ dysfunction, mental confusion, delirium, wound infection, prolonged hospital stay, and increased cost for the hospital and patient [10-14].

Factors that may involve hypoventilation, diffusion hypoxia, V/Q imbalance, atelectasis, and increased oxygen demand are respiratory depressants (opioids, anesthetic drugs), shock, prolonged surgery, pre-existing cardiovascular diseases, respiratory disease, positioning, acute trauma, ASA level, age, and obesity [15-18]. This study aims to assess the incidence and factors associated with early postoperative hypoxemia among surgical procedures.

Methods

Study design, area, period and population

This prospective observational study was conducted in Debre Tabor Comprehensive Specialized hospital from February 01, 2020 to June 30, 2020. Debre Tabor Comprehensive Specialized Hospital is a public Hospital established in 1934 and located in South Gondar Zone, Amara region 667 km NorthWest of Addis Ababa, the capital city of Ethiopia. The climate of this town is warm-moderate and has a latitude and longitude of 11051’N3801’E/11.8500N 38.0170E with an elevation of 2,706 meters above sea level [19]. The Hospital has three surgical operation rooms, two orthopedic operation rooms, and two gynecologic and obstetrics operation rooms. The study population of our study consists of patients who underwent surgery under general anesthesia and met the inclusion criteria.

Inclusive criteria

Patients older than 18 years undergoing surgery under general anesthesia, and patients who were extubated on the operation table were included.

Exclusive criteria

Preoperative hypoxemia; patients entering the surgical unit having had oxygen therapy; intraoperative use of methylene blue; problems in taking an SPO2 reading; patients who were transferred to intensive care unit (ICU) were excluded.

Sample size determination and sampling technique

The sample size was calculated by taking the following assumption; the proportion of early postoperative hypoxemia is 50%, confidence interval of 95%, and margin of error of 0.05. The sample size was determined using the following single population proportion formula:

whereas: n=sample size; z=confidence interval (1.96); p=estimated prevalence (0.5); d=margin of sampling error to be tolerated (0.05). By adding 10% of sample size (non-respondent rate) final sample size was 385+39 = 424. A non-probable convenient sampling technique was used.

Data collection

Data were collected using a structured questionnaire and the collection of information commenced from the start of the transfer of the patient from the operation theatre and continued to the first 20 min in the post anesthesia care unit (PACU). As patient returned to the recovery room, the arterial oxygen, saturation level was measured immediately with pulse oximetry then every 5 min till 20 min. A questionnaire was developed from different bibliographic research [2,20,21], which includes demographic and patient-related data, surgery-related data, and anesthetic-related data. Demographic and patient-related characteristics: age, sex, ASA, BMI, smoking status, history of the medical disease; surgery-related characteristics: Type of surgery, the urgency of the procedure, positioning, the estimated blood loss, necessity of blood transfusion; anesthetic-related characteristics: premedication, SPO2 reading before induction, IV induction agent, muscle relaxant, use of reversal (neostigmine), analgesics used, the total amount of intraoperative fluid, time from intubation to extubation, oxygen therapy during patient transfer and/or PACU, and sedation score.

Data quality assurance

After training, the questionnaire was given to data collectors, and data was properly collected and filled in the prepared format. The principal investigator will supervise the data collectors and check for the completeness of the data daily. To ensure the quality of data, a pre-test of the questionnaire was done on surgical patients at Felege Hiwot Referral hospital.

Data analysis

After completion of data collection, the data were manually checked for errors; coded, and entered into STATATM v. 16 for analysis. Descriptive statistics were used to summarize data. Bivariate and multivariate analyses were used to check the association between the independent variables with the dependent variable (early postoperative hypoxemia). Independent variables were analyzed by using binary logistic regression with the dependent variable and those with a p of ≤0.2 were fitted to a multivariate analysis. In multivariate analysis, a p of less than or equals to 0.05 was used as a cut of point for the presence of association. Tables were used to display the results of this study.

Operational definition

Early postoperative hypoxemia: the appearance of hypoxemia (SpO2 <90%) during the transfer of the patient from operation theatre and within 20 min postoperatively in the PACU [1]. Hypoxemia was categorized as mild (90% > SpO2 ≥86%), moderate (86% >SpO2 ≥ 81%), and severe (SpO2 <81%) [1].

Results

A total of 424 patients were included in this study with a response rate of 100 %. The majority (72.9%) of the study participants were between the ages of 18 to 65, and most of the study participants were ASA physical status I (46%) (Table 1).

Table 1.

Demographic and patient-related characteristics of the study participants at Debre Tabor Comprehensive Specialized Hospital 2020/2021 (n=424).

Variables	Hypoxemia		Total, n (%)
	Yes, n (%)	No, n (%)
Age (years)
18-65	139 (32.8)	170 (40.1)	309 (72.9)
>65	55 (12.9)	60 (14.2)	115 (27.1)
Sex
Female	92 (21.7)	10 7(25.2)	199 (46.9)
Male	102 (24.1)	123 (29.0)	225 (53.1)
BMI (kg/m2)
18.5-24.9	105 (24.8)	157 (37.0)	262 (61.8)
25-29.9	26 (6.1)	17 (4.0)	43 (10.1)
30-39.9	41 (9.7)	24 (5.7)	65 (15.3)
<18	22 (5.2)	32 (7.5)	54 (12.7)
ASA physical status
ASA I	91 (21.5)	104 (24.5)	195 (46.0)
ASA II	73 (17.2)	102 (24.1)	175 (41.3)
ASA III
Smoking	30 (7.1)	24 (5.7)	54 (12.7)
Yes	30 (7.1)	18 (4.2)	48 (11.3)
No	164 (38.7)	212 (50.0)	376 (88.7)
Presence of coexisting disease
Had no chronic disease	108 (25.5)	186 (43.9)	294 (69.4)
Hypertension	42 (9.9)	30 (7.1)	72(17.0)
Bronchial asthma	44 (10.4)	14 (3.3)	58(13.7)

Surgery-related characteristics of the study participants

Among the various types of surgeries, the highest proportion of early postoperative hypoxemia occurred in patients who underwent abdominal surgery (20.3%). Early postoperative hypoxemia was found to be more common in emergency surgeries compared with elective surgeries (23.1% vs 22.6% - 98/166 vs 96/258) (Table 2).

Table 2.

Surgery-related characteristics of the study participants at Debre Tabor Comprehensive Specialized Hospital 2020/2021 (n=424).

Variables	Hypoxemia		Total, n (%)
	Yes, n (%)	No, n (%)
Urgency of the procedure
Elective	96 (22.6)	16 2(38.2)	258 (60.8)
Emergency	98 (23.1)	68 (16.0)	166 (39.2)
Type of surgery
Abdominal	86 (20.3)	116 (27.4)	202 (47.6)
Thyroidectomy	36 (8.5)	48 (11.3)	84 (19.8)
Orthopedic	24 (5.7)	18 (4.2)	42 (9.9)
Gynecologic and Obstetric	48 (11.3)	48 (11.3)	96 (22.6)
Positioning
Supine	147 (34.7)	185 (43.6)	332 (78.3)
Lithotomy	47 (11.1)	45 (10.6)	92 (21.7)
Estimated blood loss (ml)
<500	79 (18.6)	120 (28.3)	199 (46.9)
500-1000	66 (15.6)	75 (17.7)	141 (33.3)
1000-1500	30 (7.1)	35 (8.3)	65 (15.3)
1500-2000	16 (3.8)	3 (0.7)	19 (4.5)
Necessity of transfusion
Yes	17 (4.0)	15 (3.5)	32 (7.5)
No	177 (41.7)	215 (50.7)	392 (92.5)

Anesthetic-related characteristics of the study participants

Among the various types of surgeries, the highest proportion of early postoperative hypoxemia occurred in patients who underwent abdominal surgery (20.3%). Early postoperative hypoxemia was found to be more common in patients who had no oxygen supplementation compared with patients on oxygen therapy during patient transfer and/or in the PACU (23.1% vs 22.6% - 98/166 vs 96/258). Early postoperative hypoxemia was common in patients who had lower oxygen saturation before induction (SPO2 <95%) as compared to patients who had higher oxygen saturations before induction (SPO2 ≥95%) (4.3% vs 41.5% - 18/56 vs 176/368) (Table 3).

Table 3.

Anesthetic-related characteristics of the study participants at Debre Tabor Comprehensive Specialized Hospital 2020/2021 (n=424).

Variables	Hypoxemia		Total, n (%)
	Yes, n (%)	No, n (%)
Premedication
Neither	103 (24.3)	122 (28.8)	225 (53.1)
Opioids	79 (18.6)	85 (20.0)	164 (38.7)
Benzodiazepines	12 (2.8)	23 (5.4)	35 (8.3)
SPO2 (%) reading before induction
≥95	17 6(41.5)	192 (45.3)	368 (86.8)
<95	18 (4.3)	38 (8.9)	56 (13.2)
Intravenous induction agent
Ketamine	24 (5.7)	30 (7.1)	54 (12.7)
Propofol	51 (12.0)	56 (13.2)	107 (25.2)
Ketofol	66 (15.6)	59 (13.9)	125 (29.5)
Thiopental	53 (12.5)	85 (20.0)	138 (32.5)
Muscle relaxant used
Suxamethonium	6 (1.4)	10 (2.4)	16 (3.8)
Vecuronium	11 (2.6)	18 (4.2)	29 (6.8)
Suxamethonium and vecuronium	146 (34.4)	177 (41.7)	323 (76.2)
Not used	31 (7.3)	25 (5.9)	56 (13.2)
Number of intubation attempts
Single	128 (30.2)	182 (42.9)	310 (73.1)
Multiple	66 (15.6)	48 (11.3)	114 (26.9)
Neostigmine used
Yes	157 (37.0)	188 (44.3)	345 (81.4)
No	37 (8.7)	42 (9.9)	79 (18.6)
Intraoperative analgesic used
Diclofenac	13 (3.1)	18 (4.2)	31 (7.3)
Pethidine	107 (25.2)	113 (26.7)	220 (51.9)
Diclofenac and pethidine	749 (17.5)	99 (23.3)	177 (41.7)
Intraoperative fluid used (ml)
<1000	24 (5.7)	23 (5.4)	47 (11.1)
1000-2000	25 (5.9)	36 (8.5)	61 (14.4)
2000-3000	84 (19.8)	135 (31.8)	219 (51.7)
>3000	61 (14.4)	36 (8.5)	97 (22.9)
O2 therapy
Yes	18 (4.2)	44 (10.4)	62 (14.6)
No	176 (41.5)	186 (43.9)	362 (85.4)
Shiver in PACU
Yes	47(11.1)	45 (10.6)	92 (21.7)
No	147 (34.7)	185 (43.6)	332 (78.3)
Sedation score in PACU
0	145 (34.2)	172 (40.6)	317 (74.8)
1	29 (6.8)	39 (8.8)	68 (16.0)
2	10 (2.4)	11 (2.6)	21 (5.0)
3	10 (2.4)	8 (1.9)	18 (4.2)
Time from intubation to extubation (min)
<60	30 (7.1)	42 (9.9)	72 (16.9)
≥60	164 (38.7)	188 (44.4)	352 (83.0)

PACU, post anesthesia care unit.

Prevalence and factors associated with early postoperative hypoxemia of the study participants

Early postoperative hypoxemia was found to be in 149 patients (45.8%). In this study, obesity was significantly associated with postoperative hypoxemia. Hypoxemia was mild in 80% of these cases (n = 155); moderate (86 N SpO2 ≥81%) in 12% of cases (n=23); and severe in 8% of cases (n=16). Patients who had a BMI of 25-29.9 kg/m2 were 2.6 times more likely to develop postoperative hypoxemia than normal patients (AOR = 2.588, 95% CI: 1.269, 5.276). Current smokers were 2.2 times more likely to develop postoperative hypoxemia compared with non-smokers (AOR = 2.192, 95% CI: 1.095, 4.391). Emergency surgical procedures (AOR = 3.193), 95% CI: 2.013, 5.064), patients who had a history of bronchial asthma (AOR = 7.519, 95% CI: 3.706, 15.256), hypertensive patients (AOR = 3.334, 95% CI: 1.835, 6.056), patients who had lower oxygen saturation before induction (SPO2 <95%) before induction (AOR = 3.011, 95% CI: 1.535, 5.908), and the absence of oxygen therapy during the period of transfer of the patient from operation theatre and/or in the PACU (AOR = 2.559, 95% CI: 1.310, 5.001) were more likely to develop early postoperative hypoxemia than their counterparts (Table 4).

Table 4.

Factors associated with early postoperative hypoxemia of the study participants at Debre Tabor Comprehensive Specialized Hospital 2020/2021 (n=424).

Variables	Hypoxemia		COR (95% CI)	AOR (95% CI)	p
	Yes	No
BMI (kg/m2)
18.5-24.9	105	157	1	1
25-29.9	26	17	2.287 (1.183, 4.422)	2.588 (1.269, 5.276)	0.009
30-39.9	41	24	2.554 (1.458, 4.477)	2.411 (1.278, 4.548)	0.007
<18	22	32	1.028 (0.566, 1.866)	1.165 (0.589, 2.304)	0.662
Smoking
Yes	30	18	2.154 (1.160, 4.001)	2.192 (1.095, 4.391)	0.027
No	164	212	1	1
Urgency
Elective	96	162	1	1
Emergency	98	68	2.432 (1.631, 3.625)	3.193 (2.013, 5.064)	<0.0001
Having chronic disease
Neither	108	186	1	1
Hypertension	42	30	2.411 (1.426, 4.076)	3.334 (1.835, 6.056)	<0.0001
Bronchial asthma	44	14	5.413 (2.836, 10.336)	7.519 (3.706, 15.256)	<0.0001
SPO2 (%) reading before induction
≥95	176	192	1	1
<95	18	38	1.935 (1.065, 3.515)	3.011 (1.535, 5.908)	0.001
O2 therapy
Yes	18	44	1	1
No	176	186	2.313 (1.288, 4.155)	2.55 9(1.310, 5.001)	0.006

Discussion

Early postoperative hypoxemia was found to be in 149 patients (45.8%). In this study, obesity was significantly associated with postoperative hypoxemia. Patients who had a BMI of 25-29.9 kg/m2 were 2.6 times more likely to develop postoperative hypoxemia than normal patients (AOR = 2.588, 95% CI: 1.269, 5.276). Despite the recent advances in anesthesia and surgical care, early postoperative hypoxemia is still a common problem after general anesthesia. The identification of factors associated with an increased occurrence of early postoperative hypoxemia could help alert staff to hypoxemia risk patients. In this study, the overall incidence of early postoperative hypoxemia was found to be 45.8%. In line with our result, a study conducted in Canada by Denise et al. showed that 41% of patients had at least one episode of hypoxemia during their PACU stay following surgery [22]. In contrary to our finding studies showed a lower incidence of early postoperative hypoxemia such as studies conducted in Japan by Ishikawa et al. [23] (12.7%), a study done in Brazil by Filho et al. [24] (24.1%), and a study in the USA by Ramachandran et al. 10.7-22% [22]. This discrepancy might be due to a variation in study settings where the above developed countries could have a better perioperative patient care as compared to the developing ones like Ethiopia. The incidence of hypoxemia in studies done in Ethiopia was 22.7% [25] and 26.7% [20] in Tikur Anbesa Comprehensive Specialized hospital and in University of Gondar hospital, respectively. In this study, we identified that patients with a BMI >25 kg/m2, having a chronic disease (hypertension and bronchial asthma), current smokers, with lower oxygen saturation before induction (SpO2 <95%), emergency surgery, and the absence of oxygen therapy during the period of transfer of the patient from operation theatre and/or in the PACU were all significantly associated with increased hypoxemia in the early postoperative period. This finding was similar to various studies, which showed that patients who had respiratory co-morbidity [25,26] and were smokers [20,23,26], patients who were obese [23,26], withlower preoperative oxygen saturation (SPO2 < 95%) [7,20,24,] were significantly associated with increased hypoxemia.

Conclusions

The findings of this study showed that obese patients, patients having a chronic disease (hypertension and asthma), current smokers, lower oxygen saturation before induction (SpO2 <95%), emergency surgery, and the absence of oxygen therapy during patient transfer and/or in the PACU were the main predictors of an increased occurrence of early postoperative hypoxemia. The identification of these factors may help to alert healthcare professionals to hypoxemia risk patients.