Perioperative outcomes in human immunodeficiency virus-infected patients

Abstract

It is estimated that a quarter of patients with HIV/AIDS undergo at least one surgical procedure in their life time. Surgical outcomes in these patients from developing countries are poorly characterized and surgeons are often concerned about poor surgical outcomes, especially when their CD4 cell counts are less than 200 cells/µl. This study evaluated the surgical outcomes of HIV-infected patients undergoing various surgical procedures over a six-year period in a large tertiary care hospital from South India. Two hundred and ninety-three patients underwent 374 surgical procedures during the study period. The median duration of HIV prior to surgery was 1.9 years (range 0–18.8 years). Two-thirds (58%) were on highly active antiretroviral therapy (HAART) at the time of surgery with the median duration of this treatment being 38 months (n = 194). About one-third (35%) of surgical procedures were performed as an emergency. Abdomino-pelvic surgeries were the most common (225, 60%). Adverse surgical outcome defined as death or post-operative infection was seen in 25 (6.6%). The post-operative infection rate was 5% (20/374). The most common of these was surgical site infection observed in nine (60%) followed by pneumonia in five patients (33%) and urinary tract infection in one patient. Day 30 mortality was 2% (n = 8) and a quarter of these were reported to be related to post-operative infectious complications. On multivariate analysis, only preoperative haemoglobin of less than 10 g/dl was significantly associated with a poor surgical outcome. HIV-related parameters such as CD4 cell counts, duration of HIV infection and HAART regimen did not seem to contribute towards an adverse surgical outcome.

Keywords

Perioperative outcomes surgery in HIV post-operative mortality post-operative infections

Background

Close to 40 million people live with HIV/AIDS in the world with India having the third largest number of people living with HIV/AIDS (PLHA).¹ The advent of effective highly active antiretroviral therapy (HAART) has changed the clinical spectrum of HIV from managing life-threatening opportunistic infections to a chronic disorder with normal life expectancy. About a quarter of patients are likely to undergo an elective or emergency surgery during their lifetime, which could be unrelated to their diagnosis of HIV/AIDS.² Surgeons are often concerned about high perioperative mortality and morbidity and poor wound healing in patients with HIV infection, especially if their CD4 cell counts are less than 200 cells/µl. This leads to a reluctance and delay in elective surgeries among these patients and a high-risk perception when emergency surgeries are performed. These concerns are due to the high surgical mortality documented in the studies done in the pre- or early HAART era from developed nations.^3–5 The high mortality was primarily attributed to opportunistic infections, organ dysfunctions and low CD4 cell counts in addition to patient-related factors such as low albumin in these patients.

However, much has changed with universal availability of effective antiretroviral therapy (ART). More than a million patients are receiving regular ART in India, with the roll-out of combination ART by the National AIDS Control Organization. The surgical outcomes among HIV-infected individuals in India have not been studied before. The data from developed countries may not be applicable to a developing country like India owing to delayed clinical presentation of disease, limited access to healthcare facilities, increased opportunistic infections and delayed initiation of HAART. Also, the unique health structure with a mix of private and public sector is different from many underdeveloped countries. Opportunistic infections are increasingly uncommon in the developed countries, while they continue to contribute to significant morbidity in India. Hence, we felt that a study aimed at evaluating the perioperative outcomes and evaluating risk factors which could predict adverse outcomes among PLHA would be appropriate.

Patients and methods

This is a retrospective single centre study carried out at a large tertiary care centre in South India. All HIV-infected patients who underwent elective and emergency surgeries from January 2010 to December 2015 at our hospital were included. Patients undergoing diagnostic biopsies were excluded. All relevant details were retrieved from patients’ inpatient and outpatient charts and electronic medical records.

Apart from basic demographic data, other patient-related data collected included the month and year of diagnosis of HIV, baseline and preoperative (within 90 days prior to surgery) CD4 cell count and HIV viral load, HAART use prior to surgery, the current and prior HAART regimen(s) and reason for change if any. The revised WHO system⁶ was used for identifying the preoperative clinical stage of patients. Surgical details including type and anatomic region of surgery along with the indication and use of any prosthesis or foreign body were collected. Routine laboratory investigations done preoperatively were also recorded.

Data regarding use and duration of pre- and post-operative antimicrobials, duration of post-operative ICU stay (if any) and total hospital stay were noted from inpatient records and reconfirmed using pharmacy records. Post-operative infective complications and status at 30 days post-surgery were noted. For patients who died, the cause of death was categorized as due to (1) HIV, (2) underlying indication for surgery, (3) infective or non-infective post-operative complication or (4) unrelated cause. These data were generated from retrospective chart reviews as most of the patients are followed up in the ART centre and mandatory regular follow-ups were part of routine post-operative care. Permission was obtained from all concerned surgical units for the use of data. The study protocol was reviewed and approved by the institutional review board and consent waiver was granted.

The primary outcome measure was the incidence of a 30-day adverse surgical outcome (post-operative infection and death). The secondary outcome measures included incidence of various post-operative infective complications, sepsis and unplanned ICU admissions. Potential risk factors as mentioned above were evaluated for their association with post-operative infection and death. For statistical analysis, Fisher’s exact or Chi square test was used for discrete variables. For continuous variables, T-test was used. All p values were two sided. Statistical analysis was performed using SPSS software (version 18).

Results

From 1 January 2010 to 31 December 2015, of the 2844 HIV-infected patients admitted, 293 patients underwent 374 surgeries. Most of our patients were middle-aged men. Median duration of HIV infection prior to surgery was 1.9 years (range 0–18.8 years). More than half of the patients (58%) were on HAART at the time of surgery with the median duration of HAART being 38 months (n = 194). Median preoperative CD4 cell count was 389 (range 9–2159) cells/mm³. Baseline CD4 cell counts available for 248 (66%) patients showed a median value of 256 (range 9–1300) cells/mm³. The commonest HAART regimen at the time of surgery was zidovudine (AZT)/lamivudine (3TC)/nevirapine (NVP) in 66 (30%) followed by tenofovir (TDF)/3TC/efavirenz (EFV) in 63 (29%) cases. Of the 215 cases who were ART exposed, 72 (33%) had undergone a change in therapy, with the most recent change in ART being due to drug toxicity in 53 (74%), treatment failure in 12 (17%) and other reasons in 7 (9%). The preoperative WHO clinical stage was distributed in the cohort as follows: stage I (51%), II (8%), III (21.5%) and IV (20%) suggesting that about a quarter of patients had advanced HIV infection at the time of surgery.

Preoperative viral load was available only in nine (2%) cases ranging from 66 to 587,745 viral copies per millilitre (level of detection <40 copies/µl). Median preoperative haemoglobin (n = 318) and albumin (n = 156) were 11.6 (range 2.3–18.5) and 3.5 (1.6–5.2) g/dl, respectively. These characteristics are summarized in Table 1.

Table 1.

Baseline characteristics of patients.

Characteristics	Numbers (%)
Mean age of patients^@	40.05 (12.9)
Mean baseline CD₄ cell count^@	305 (230.9)
Mean pre-op CD4^@	406 (297.1)
Mean pre-op albumin^@	3.48 (0.84)
Mean pre-op haemoglobin^@	11.6 (4.7)
Mean pre-op total count^@	11,401 (31,697)
Mean pre-op platelet^@	244,807 (116,958)
Gender	Male	252 (67)
Gender	Female	122 (33)
Duration of HIV prior to surgery	<1 year	166 (44.4)
	1–5 years	95 (25.4)
	>5 year	99 (26.5)
Previous ART use	Yes	215 (57.5)
Previous ART use	No	153 (40.9)
Duration of prior ART use	Up to 12 months	58 (26.9)
Duration of prior ART use	>12 months	136 (74.1)
Preoperative CD₄ cell count	Up to 200 cells	72 (27.4)
Preoperative CD₄ cell count	>200 cells	189 (72.4)
Preoperative haemoglobin	Up to 10 g	99 (26.3)
Preoperative haemoglobin	>10 g	278 (73.7)
Preoperative albumin	Up to 2.5	28 (7.4)
Preoperative albumin	>2.5	349 (92.6)

ART: antiretroviral therapy.

Mean (SD)^@.

Surgical details

Of the 374 surgeries, 134 (35%) were emergency surgeries. Two hundred and ninety-two (78%) were open surgeries and a majority involved the abdomino-pelvic region (225, 60%). Debridement of infected area or drainage of abscess was the commonest indication (101, 27%) followed by obstetric indications (49, 13%). A significant proportion of our patients (24.9%) had foreign material such as prostheses or vascular grafts implanted during the surgery. The surgical details are summarized in Table 2.

Table 2.

Surgical characteristics of patients.

Characteristics	Category	N (%)
Type of surgery	Emergency	134 (35.8)
Type of surgery	Elective	240 (64.2)
Type of surgery	Open surgery	292 (78.1)
Type of surgery	Minimally invasive	82 (21.9)
Anatomic regionof surgery	Abdominal	225 (59.7)
Anatomic regionof surgery	Other	152 (40.3)
Indication for surgery	Infection, traumaand cancer	187 (50)
Indication for surgery	Obstetric and others	187 (50)
Prosthesis use	None	281 (75.1)
	Metallic prosthesis	22 (5.9)
	Stent, vasculargraft and others	71 (19)
Type of anaesthesia	General	191 (51.1)
Type of anaesthesia	Spinal and local	173 (48.9)

Table 3.

Factors associated with adverse surgical outcomes among HIV-infected patients.

Variable	Category	Adverse surgical outcome n (%)	Good surgical outcome n (%)	OR (95%CI)	P Value
Age	More than 40 years	11 (5.39)	193 (94.6)	1.57 (0.69–3.5)	0.27
Age	Up to 40 years	14 (8.23)	156 (91.76)	1.57 (0.69–3.5)	0.27
Gender	Male	16 (6.34)	236 (93.6)	0.85 (0.36–1.98)	0.7
Gender	Female	9 (7.3)	113 (92.6)	0.85 (0.36–1.98)	0.7
Type of surgery	Emergency	12 (8.9)	122 (91)	1.7 (0.7–3.8)	0.18
Type of surgery	Elective	13 (5.41)	227 (94.5)	1.7 (0.7–3.8)	0.18
Surgery type	Open	22 (7.5)	270 (92.4)	2.14 (0.6–7.3)	0.21
Surgery type	Minimally invasive	3 (3.6)	79 (96.3)	2.14 (0.6–7.3)	0.21
Anatomical region	Abdominal	21 (9.3)	204 (90.6)	3.7 (1.2–11)	0.01
Anatomical region	Other	4 (2.6)	145 (97.3)	3.7 (1.2–11)	0.01
Indication for surgery	Trauma, infection and cancer	15 (8.02)	172 (91.9)	1.54 (0.67–3.5)	0.3
Indication for surgery	Obstetric and others	10 (5.34)	177 (94.6)	1.54 (0.67–3.5)	0.3
Type of anaesthesia	General	21 (11)	170 (89)	5.5 (1.8–16)	0.001
Type of anaesthesia	Spinal and others	4 (2.1)	179 (97.8)	5.5 (1.8–16)	0.001
ART use	ART naïve	9 (5.8)	144 (94.1)	0.8 (0.35–1.9)	0.67
ART use	ART experienced	15 (6.9)	200 (93)	0.8 (0.35–1.9)	0.67
Pre-op CD4	Up to 200	8 (11.1)	64 (88.8)	0.44 (0.16–1.1)	0.09
Pre-op CD4	>200	10 (5.2)	179 (94.7)	0.44 (0.16–1.1)	0.09
Pre-op Alb	Up to 2.5	6 (21.4)	22 (78.6)	4 (1.2–12.9)	0.01
Pre-op Alb	>2.5	8 (6.25)	120 (93.75)	4 (1.2–12.9)	0.01
Pre-op Hb	Up to 10 g	14 (14.1)	85 (85.8)	3 (1.5–9.1)	0.002
Pre-op Hb	>10 g	9 (4.1)	208 (95.8)	3 (1.5–9.1)	0.002
ICU admission	Yes	9 (37.5)	15 (62.5)	12.5 (4.7–32.9)	<0.001
ICU admission	No	16 (4.5)	334 (95.4)	12.5 (4.7–32.9)	<0.001
Empirical antifungal use	Yes	6 (37.5)	10 (62.5)	10.7 (3.5–32.5)	<0.001
Empirical antifungal use	No	19 (5.3)	339 (94.6)	10.7 (3.5–32.5)	<0.001

CI: confidence interval; ICU: intensive care unit; OR: odds ratio; Alb: albumin; Hb: haemoglobin.

Table 4.

Multivariate analysis.

Variable	OR	95% CI	P value
Anatomical region of surgery	1.08	0.30–3.85	0.907
Anaesthesia category	0.49	0.14–1.73	0.269
Preoperative albumin	2.48	0.66–9.27	0.178
Preoperative haemoglobin	4.13	1.04–16.37	0.044*
Antifungal therapy	0.20	0.03–1.20	0.078
ICU stay	4.17	0.86–20.21	0.076

CI: confidence interval; ICU: intensive care unit; OR: odds ratio.*Statistically significant

Post-operative outcomes

A good surgical outcome was seen in 349 (93.4%) patients with a mean preoperative CD4 cell count of 419 cells/µl. Adverse surgical outcome as defined by death or post-operative infection was seen in 25/374 (6.6%) patients. Post-operative infection rate was 5% (n = 20). The focus of most of the post-operative infection was the surgical site in nine (45%) followed by urinary tract in six patients (30%) and pneumonia in five patients (25%). Blood cultures were sent in 38 cases of which five (13%) were positive. Twenty-five cases (6%) required ICU admission. Four patients (1%) required prolonged post-operative ventilatory support. The median duration of ICU stay was 5.5 (range 1–60) days. Eight patients (2%) had died by day 30 post-surgery and four who were discharged against medical advice were lost to follow-up. Among the eight deaths, six (75%) were related to the underlying indication for surgery and two (25%) were related to post-operative infective complications.

Of the factors analysed, general anaesthesia (p < 0·001), abdominal surgeries (p = 0·01), low preoperative albumin (up to 2.5 g/dl) (p = 0·01), low preoperative haemoglobin (up to 10 g/dl) (p = 0.01), ICU stay (p < 0.001) and empirical antifungal use (p < 0·001) were associated with increased risk of poor surgical outcome in univariate analysis (Table 3). Multivariate analysis showed that only a preoperative haemoglobin up to 10 g/dl contributed to an adverse outcome (p = 0.044) (Table 4).

Discussion

Our study is one of the largest analyses of surgical outcomes for HIV-infected patients from a developing country in the modern combination ART era. Our study concludes that the perioperative mortality and infectious morbidity in our patients is considerably low. The risk factors for poor outcome in our study were reflective of general health status of the patients such as low haemoglobin. Factors suggestive of active or uncontrolled HIV infection including low CD4 cell counts and not being on ART before the surgery did not contribute to a poor outcome, despite being important in predicting long-term outcomes of the patient. The importance of our study results rests on the large number of patients and all types of surgeries being represented. The perioperative infectious morbidity and death in our study is quite similar to surgical outcomes in HIV-infected patients from developed countries.^7,8

The studies evaluating perioperative outcomes in HIV infection are mostly from developed nations with few reports from underdeveloped countries. In a recent large retrospective study involving nationwide data from a developed nation with procedure-matched uninfected comparators, lower CD4 cell counts, advanced age and hypoalbuminaemia were associated with increased mortality.⁷ Other studies have shown similar outcomes with a slightly higher risk of post-operative pneumonia, with no excessive perioperative morbidity attributable to HIV infection.^8–10 Studies from Africa have also showed similar conflicting results.^11–13 A recent large multinational prospective cohort study also did not show poor outcomes among HIV-infected patients.¹⁴ Studies have also differed in the patient populations, with some studies focussing solely on cardiac surgery,¹¹ anorectal surgeries,¹² orthopaedic surgeries,¹³ etc. However, the common surgeries included in the above analyses closely reflect the general population while in our study a large number of surgeries were done to address infectious morbidity. The huge burden of tuberculosis and other opportunistic infections likely contribute to a significant number of surgeries in a resource-limited setting like India.

The incidence of post-operative surgical site infection depends on patient factors like complexity of the underlying indication for surgery, associated co-morbidities, nutritional status of the patient, hyperglycaemia, preoperative site of infection, smoking, hypothermia and immune status of the patient. It would be intuitive to assume that a low CD4 cell count as a result of HIV infection would predispose to an increased risk of opportunistic infection but whether this would translate to an increased risk of post-operative infection is controversial. In HIV infection, there is a selective early depletion of the cell-mediated immunity, while in advanced untreated disease with low CD4 cell counts there is a significant risk of morbidity and death due to opportunistic infections. Concurring with the above, studies reporting surgical outcomes in the pre- and early HAART era reported excess morbidity and mortality.^3–5 A majority of the studies shows that low CD4 cell counts are associated with increased risk of post-operative complications and death.^2,5,7 In most of the studies, the cut-off for low CD4 cell count was taken to be 200 cells/µl. Our study did not confirm an association of low CD4 cell count with risk of post-operative complications. One potential explanation could be that patients with low baseline CD4 cell counts may not show significant increase in their CD4 cell counts even after the initiation of HAART; however, their low CD4 cell counts may not put them at a risk similar to those patients who are HAART naive.

The studies evaluating post-operative complications have evaluated wound infections,^2,5,14 all wound complications (including non-infective complications like healing)^5,7 and others evaluating all post-operative complications.^18–20 Some studies including high-risk conditions like traumatic orthopaedic wounds report a high rate of wound infections, as high as 39%.¹⁷ Studies which have included an unselected cohort of HIV patients have reported an overall post-operative complication rate of about 10%, which was similar to cohorts of non-HIV patients.⁸ We studied all infective complications and found the rate to be 4%. The pattern of infective complications is also similar to that observed in other similar studies, with surgical site infections and pneumonia being the common post-operative sites of infection.⁸

The predictive value of patient factors like low preoperative haemoglobin for surgical outcomes noted in our study has been reported by other studies also.^7,16 However, given the retrospective nature of other studies and ours, it would be difficult to comment whether an improvement in preoperative haemoglobin or albumin would lead to any clinically significant improvement in the outcomes. It appears likely that these adverse prognostic factors are related to the underlying illness (like infection, chronic inflammation, etc.) and HIV rather than a pure nutritional deficiency.

In addition to inherent limitations of retrospective analyses, our study did not have a matched control group to accurately study the impact of HIV infection on outcomes. Although the surgical data were available for all patients, some other data like WHO clinical stage, preoperative CD4 cell count and even albumin were missing in some patients. Also, data for plasma viral RNA were present for so few patients that it could not be assessed.

To conclude, we report a large study of the surgical outcomes in HIV patients from a developing country. Contrary to the general belief of poor outcomes of patients in developing countries, we found a low rate of post-operative infections and death in our cohort of patients, which is comparable or lower than many other studies reported from the developed nations. Surgical morbidity and mortality in HIV-positive patients appears to be comparable to HIV-negative patients. We also identified preoperative risk factors which could help in identifying patients at an increased risk of complications. Addressing these risk factors might help to further improve the outcomes in this cohort of patients, thus eliminating concerns regarding adverse surgical outcomes.

Footnotes

Authors’ contributions

DS – Contributed to literature search, data collection, data interpretation and writing of the manuscript.

AM – Contributed to data collection, writing and revision of the manuscript.

NGPZ – Did the data analysis.

LRR – Contributed to data collection.

PR – Concept, study design, data interpretation and critical revision of the manuscript.

Acknowledgements

We gratefully acknowledge all the surgical departments at Christian Medical College Vellore who permitted us to analyse the surgical outcomes in their patients. We could not have done without their consent and cooperation.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

World Health Organization. HIV AIDS fact sheet, http://www.who.int/mediacentre/factsheets/fs360/en (accessed 6 April 2017).

Dua

Wajed

Winslet

MC.

Impact of HIV and AIDS on surgical practice. Ann R Coll Surg Engl 2007; 89: 354–358.

Semprini

Castagna

Ravizza

et al . The incidence of complications after caesarean section in 156 HIV-positive women. AIDS 1995; 9: 913–917.

Grubert

Reindell

Kästner

et al . Complications after caesarean section in HIV-1-infected women not taking antiretroviral treatment. Lancet 1999; 354: 1612–1613.

Albaran

Webber

Steffes

CP.

CD4 cell counts as a prognostic factor of major abdominal surgery in patients infected with the human immunodeficiency virus. Arch Surg 1998; 133: 626–631.

WHO case definitions of HIV for surveillance and revised clinical staging and immunological classification of HIV-related disease in adults and children. Geneva: World Health Organization, 2007.

King

Perkal

Rosenthal

et al . Thirty-day postoperative mortality among individuals with HIV infection receiving antiretroviral therapy and procedure-matched, uninfected comparators. JAMA Surg 2015; 150: 343–351.

Horberg

Hurley

Klein

et al . Surgical outcomes in human immunodeficiency virus-infected patients in the era of highly active antiretroviral therapy. Arch Surg 2006; 141: 1238–1245.

Filsoufi

Salzberg

Harbou

et al . Excellent outcomes of cardiac surgery in patients infected with HIV in the current era. Clin Infect Dis 2006; 43: 532–536.

10.

Watts

Lambert

Stiehm

et al . Complications according to mode of delivery among human immunodeficiency virus-infected women with CD4 lymphocyte counts of < or = 500/microL. Am J Obstet Gynecol 2000; 183: 100–107.

11.

Green

Kong

Odendaal

et al . The effect of HIV status on clinical outcomes of surgical sepsis in KwaZulu-Natal Province, South Africa. S Afr Med J 2017; 107: 702–705.

12.

Cacala

Mafana

Thomson

et al . Prevalence of HIV status and CD4 counts in a surgical cohort: their relationship to clinical outcome. Ann R Coll Surg Engl 2006; 88: 46–51.

13.

Chichom-Mefire

Azabji-Kenfack

Atashili

CD4 count is still a valid indicator of outcome in HIV-infected patients undergoing major abdominal surgery in the era of highly active antiretroviral therapy. World J Surg 2015; 39: 1692–1699.

14.

Biccard

Madiba

Kluyts

et al.

Perioperative patient outcomes in the African Surgical Outcomes Study: a 7-day prospective observational cohort study. Lancet. Epub ahead of print 3 January 2018. doi: 10.1016/S0140-6736(18)30001-1.

15.

Polanco

Itagaki

Chiang

et al . Changing prevalence, profile, and outcomes of patients with HIV undergoing cardiac surgery in the United States. Am Heart J 2014; 167: 363–368.

16.

Lord

RV.

Anorectal surgery in patients infected with human immunodeficiency virus: factors associated with delayed wound healing. Ann Surg 1997; 226: 92–99.

17.

Abalo

Patassi

James

et al . Risk factors for surgical wound infection in HIV-positive patients undergoing surgery for orthopaedic trauma. J Orthop Surg 2010; 18: 224–227.

18.

Hajek

Novak

Zikmundova

Surgery-related death, complicated wounds, and anastomosis healing in HIV-positive patients with considerable immune deficit: assumption and reality. Int Surg 2009; 94: 228–236.

19.

Davis

Corless

Gazzard

et al . Increased risk of wound complications and poor healing following laparotomy in HIV-seropositive and AIDS patients. Dig Surg 1999; 16: 60–67.

20.

Yii

Saunder

Scott

DF.

Abdominal surgery in HIV/AIDS patients: indications, operative management, pathology and outcome. Aust NZ J Surg 1995; 65: 320–326.

Perioperative outcomes in human immunodeficiency virus-infected patients – the PRO HIV study