Secondary Logo

Journal Logo

Global Health: Original Clinical Research Report

Retrospective Comparison of 2 Management Strategies for Perioperative Malaria Episodes in Pediatric Patients in a Limited-Resource Setting

Roark, Gary L. MD

Author Information
doi: 10.1213/ANE.0000000000004186
  • Free



  • Question: What practices can help to minimize risks of perioperative fevers and clinical malaria in a limited resource setting where malaria prevalence is high?
  • Findings: Malaria treatment 3–7 days before elective surgery, with artemisinin-based therapy obtained from a reliable supply chain, may help reduce the risks of malaria episodes around the time of surgery.
  • Meaning: While the potential for cryptogenic malaria relapses and counterfeit medications may present additional problems, malaria treatment 3–7 days before elective surgeries may reduce the risks of febrile episodes and clinical malaria developing in the perioperative time frame.

Malaria is a leading cause of morbidity and mortality throughout the world, contributing to significant health effects on adults and children. Almost half of the world’s population lives in malaria-endemic regions.1 Mortality rates range from half to over a million deaths per year. Over 90% of deaths from malaria occur in sub-Saharan Africa. A disproportionate percentage of these deaths are in children <5 years of age in sub-Saharan Africa. Plasmodiumfalciparum contributes to the vast majority of this disease burden.2,3 In 2017, Niger’s population of 21 million had an estimated 7.7 million cases of malaria (95% CI, 4.1–12.8 million).2 When working in an endemic region, most anesthesiologists and surgeons from nonendemic areas of the world have minimal training or experience with one of the most important diseases affecting the populations they serve. Very little has been written about the practical management of malaria in the context of humanitarian surgical assistance projects in endemic regions of the world.1 Our study evaluated the efficacy of several approaches to managing perioperative malaria in this retrospective review.


We noticed a high incidence of perioperative fevers and malaria in a group of children having elective surgery to repair cleft lips and palates during our first 1-week surgical camp at a children’s hospital in Niamey, Niger. These children had all been tested previously for malaria by light microscopy. When positive, they were administered locally sourced artemether/lumefantrine malaria treatment 3–4 weeks before their elective surgeries. Despite treatment, several exhibited signs and symptoms of malaria around the time of surgery. We reevaluated our perioperative management strategies for 2 subsequent surgical camps for similar pediatric patients.


Approval for this retrospective review of patient charts was obtained from CURE Hôpital des Enfants au Niger Institutional Review Board and Ethics Panel. (CURE International is a nonprofit charitable organization that provides specialized medical care in the developing world.) Specific attention was given to avoid violating any ethnic, tribal, cultural, or religious practices of our patient populations. We retrospectively compared outcomes of fevers and malaria in 16 pediatric patients at our first surgical camp (group A, n = 16), with 2 subsequent surgical camps consisting of 21 and 32 additional pediatric patients having similar cleft lip/palate repairs (group B, n = 53). A surgical “camp” was typically 7–10 days long. Malaria was diagnosed in any child with an axillary temperature of ≥37.5°C and smear-positive light microscopy in group A. In group B, we used a similar axillary fever of ≥37.5°C with a positive rapid diagnostic test for the diagnosis of malaria. Light microscopy and/or rapid diagnostic tests along with a documented fever are the standard criteria used to diagnose malaria in Niger and around the world. Malaria may present with just fever but can be accompanied by malaise, headache, abdominal pain, nausea or vomiting, chills, and rigors and includes a wide spectrum of presentations, ranging from seizures to more severe and often fatal neurological sequelae when microvascular complications arise from falciparum malaria (end-organ complications due to “cerebral malaria”). Standard treatment options for malaria in Niger include artemether/lumefantrine for uncomplicated malaria and parenteral artesunate/quinine for complicated malaria.2

Patients with unrepaired cleft lips and cleft palates were recruited from all across Niger and brought with a family member to the home base of a charitable foundation located in Zinder, Niger. Written informed consent for treatment was obtained from all patient families. All 16 pediatric patients in group A were tested for malaria in Zinder using light microscopy 3–4 weeks before coming to our hospital for surgery in Niamey, 462 miles (744 km) away. Each of the 4 patients with a positive test was then administered artemether/lumefantrine (artemisinin-based combination therapy) sourced in Zinder. The antimalarial medications were purchased locally from street vendors, packaged in what appeared to be genuine labels, similar to medications obtained from reputable supply chains. Malaria parasites were diagnosed by thick smear light microscopy, the standard method at the time.

We modified our protocols after the first surgical camp due to the high number of fevers and malaria episodes during and after surgeries in group A patients and the many problems encountered as a result. We made the following changes for subsequent patients in group B:

  1. We added routine rapid diagnostic tests for malaria to include both P. falciparum and all other plasmodium species (05FK60 Standard Diagnostics Malaria Antigen Rapid Diagnostic Test; Standard Diagnostics, Kyonggi, Republic of Korea). Rapid diagnostic tests generally provide a more sensitive and specific test than routine light microscopy for malaria diagnosis, particularly in low-resource settings and with uneven microscopy stains often appearing as a positive smear. We suspected that low-grade, subclinical malaria, and especially, nonfalciparum species, were greater contributors to the malaria burden than generally suggested by the local and international medical community. Rapid diagnostic tests use simple chromatography to assess a microdrop of blood for the presence of protein antigens from Plasmodium parasites. Many tests use a histidine-rich protein (HRP-2) specifically from the P. falciparum species, or a protein antigen from all species, Plasmodium lactate dehydrogenase (pLDH) and the presence of the former indicates falciparum malaria, while presence of the latter antigen on rapid diagnostic tests is an indicator for nonfalciparum species. At this time, most rapid diagnostic tests will not be able to distinguish the various other human parasites, such as Plasmodiummalariae, Plasmodiumovale, Plasmodiumvivax, or Plasmodiumknowlesi, and are lumped into the “nonfalciparum” category. We chose to test all patients with rapid diagnostic tests, as well as light microscopy, because we did not know exactly what was causing so many febrile complications and wanted better surveillance to investigate the cause more thoroughly. We also wanted to reduce the 3- to 4-week interval between malaria treatment and surgery.
  2. We changed the source of the antimalarial medications from street vendors in Zinder to CURE hospital medications, supplied by sources with proven track records and reliable supply chains (Durbin PLC, Middlesex, United Kingdom, a reliable British wholesaler, and/or IDA Foundation, Amsterdam, the Netherlands, a not-for-profit Dutch charitable organization). While the antimalaria medications given to the patients in group A in Zinder appeared to be packaged by a reliable pharmaceutical firm, we suspected that the medications may have been counterfeit. Our concern was based partly on examination of the packaging, which was different from reputable sources obtained through reliable supply chains. We also discovered that the street vendor prices were well below the typical prices at more reputable in-country and international pharmacies (250 vs 5000 francs).
  3. All patients in group B received routine hospital-supplied artemether/lumefantrine (Novartis, Basel, Switzerland) for malaria on admission to our hospital 3–7 days before their surgery regardless of rapid diagnostic test or light microscopy results due to the elevated number of malaria episodes noted in group A. While the use of routine antimalaria treatment is not usually recommended, we had to weigh the benefits of universal treatment for pediatric patients versus the risks of complications, potentially severe, in the presence of high rates of malaria transmission in our population. Our decision to use presumptive treatment is not without precedent. A program for intermittent preventive treatment of (presumed) malaria for pregnant women is just one part of the worldwide campaign to eradicate malaria and has reduced maternal and infant mortality rates. More recently, a similar program for children <5 years of age (intermittent preventive treatment for [presumed] malaria in children) with periodic administration of antimalaria medications for presumed malaria is also currently part of the multifaceted international approach to combating malaria in this part of Western Africa.2 There appear to be no long-term problems or increased episodes of malaria for ≤1 year in patients treated presumptively with intermittent preventive treatment of (presumed) malaria for pregnant women and with intermittent preventive treatment for (presumed) malaria in children, and indeed provides better health measures.2 However, we chose to extend treatment to all pediatric patients, rather than just those <5 years of age, and we utilized artemisinin-based combination therapy rather than amodiaquine/sulfadoxine/pyrimethamine currently recommended for intermittent preventive treatment for (presumed) malaria in children due to our more readily available supply of artemisinin-based combination therapy, concerns about toxicities with amodiaquine-based therapies during anesthesia, and patient acceptance of amodiaquine-based therapies. Amodiaquine, besides causing some gastric discomfort, has been associated with dysrhythmias, hepatotoxicity (1 in 15,000), and blood dyscrasias (1 in 2200). At the time, widespread resistance to sulfadoxine was reported. Evidence currently seems to indicate that resistance to artemisinin-based combination therapy has not yet been a substantial problem in Africa, although the potential remains. Nevertheless, it is difficult to recommend universal preoperative artemisinin-based combination therapy treatment for all pediatric patients without further studies, particularly given the current World Health Organization guidelines utilizing other medications, albeit for intermittent preventive treatment for (presumed) malaria in children rather than for surgery patients and with varying drug resistance patterns and different Plasmodium strains in other regions of the world.
  4. All other treatment and clinical management protocols and care for patients in groups 1 and 2 remained unchanged. Follow-up for all patients was limited to 2 weeks after surgeries when patients returned home.

Statistical Analysis

Fisher exact test was used for statistical analysis to compare percentages of patients developing perioperative malaria in groups A and B. Clopper–Pearson method was used to estimate 95% CIs for percentage of patients having preoperative malaria and for percentage of patients having nonfalciparum malaria.


In the first group of 16 patients, only 4 had positive malaria smears in Zinder, but on arrival for preoperative testing at the CURE hospital, 9 children had positive smears (56%; 95% CI, 30%–80%), and these 9 patients had been treated with artemisinin-based combination therapies sourced from Zinder. Of the 16 pediatric patients in the first group, 6 developed signs and symptoms of clinical malaria in the perioperative period, from the start of surgery to the first 72 hours after surgery. Of these 6 patients who developed malaria, 2 developed a fever during surgery under sevoflurane anesthesia, 3 had a fever in postanesthesia care unit before discharge to the ward, and 1 developed a temperature of 37.6°C on the second day postoperatively. All 6 patients were retested using light microscopy and were positive for malaria. All were then administered artemisinin-based combination therapy from our hospital supply.

With the change in management protocols, in the 2 subsequent surgical clinics of 21 and 32 children (group B), none of these 53 children developed perioperative or postoperative fevers or malaria symptoms, although 34% (18/53) tested positive for malaria on routine preoperative rapid diagnostic test (95% CI, 22%–48%) and were subsequently treated before their surgeries with CURE hospital-supplied artemisinin-based combination therapies. None were tested during or after surgery unless they developed a fever or other symptoms clinically suspicious for malaria. See the Table below for a comparison of the rates of preoperative and postoperative malaria episodes in the 2 groups. (Comparison of 6 of 16 patients in group A developed confirmed clinical malaria versus 0 of 53 patients in group B developed clinical malaria, P < .0001 using Fisher exact test.)

Comparison of Outcomes Between 2 Cohorts

During the first 4 months of laboratory services after implementing rapid diagnostic tests, we tested 283 patients, of whom 73 were found to be positive for malaria by light microscopy and/or rapid diagnostic test. Of the 73 malarias, 24.6% were nonfalciparum malarias (95% CI, 14.7%–34.5%), and 79.5% were falciparum (95% CI, 70.2%–88.8%). Three patients were positive for both falciparum and nonfalciparum species.


Our small retrospective study found that perioperative fevers and episodes of clinical malaria may possibly be reduced by administration of antimalaria treatment 3–7 days before surgery. Also, the use of rapid diagnostic tests may assist in determining the presence of various plasmodium species, with implications for further treatment. While our change in preoperative management strategies reduced some of the complications related to malaria, 2 significant findings remain and may have contributed to the perioperative fevers and malaria in our first group: (1) that medications obtained from local sources may have been counterfeit or may have contained lower doses than stated and (2) that the incidence of nonfalciparum malarias was much higher than previous research had suggested.

Our newer diagnostic rapid diagnostic tests found that 24.6% of the malarias were nonfalciparum species, some of which may be P. vivax and/or P. ovale, which are capable of hepatic cryptogenic malaria, that is, remaining protected from standard antimalaria treatments by hiding in the liver. When we first began performing surgeries at the new hospital, we were not aware of the high incidence of nonfalciparum malarias and saw no need to address what was originally considered a rare problem. We specifically added a rapid diagnostic test that assessed for both falciparum and nonfalciparum malaria species when we observed a number of patients returning to our outpatient laboratory for repeat malaria testing several weeks after a previous treatment. We suspected hepatic cryptogenic malaria, which does not normally respond to artemisinin-based combination therapy, but requires several weeks of an 8-aminoquinoline (primaquine or the newly approved tafenoquine) for eradication. Our finding of an elevated incidence of nonfalciparum species runs counter to most published surveys of malaria in this region of Africa. Most reports typically suggest only a 0%–5% incidence of the more “benign” nonfalciparum malarias.2,3 Because our nonfalciparum malarias may well be relapses from hepatic hypnozoites, we are not able to conclude that universal treatment with artemisinin-based combination therapies was responsible for elimination of all malaria in our patients. However, it did appear to reduce the short-term risks of febrile episodes and clinical malaria around the time of surgery.

Of note, we did not address the problem of treating malaria relapses, which requires either primaquine or tafenoquine. First, we did not have a reliable test for glucose-6-phosphate dehydrogenase (G6PD) deficiency to enable us to treat cryptogenic malaria safely with either medication. Second, the risk of complications and drug interactions with these medications under general anesthesia is unknown, and with the longer treatment required for short-term surgical relief teams, we did not believe the benefits exceeded the risks before the conduct of anesthesia.

In contrast, World Health Organization and other stakeholders do not recommend universal treatment with antimalaria medicines for just febrile episodes without a good diagnosis, because it may be one contributor to the development of drug resistance.2,4 While we appreciate the need to address longer-term issues of drug resistance, our study was based on surgical teams arriving for surgical camps for 7–10 days and the immediate need to manage surgically treatable disabilities along with the common problem of malaria with risks of potentially severe complications due to malaria in these pediatric patients. The use of better diagnostic techniques such as rapid diagnostic tests may, well in the future, preclude the need for universal treatment and allow treatment of only those patients with a parasitological rather than presumptive diagnosis of malaria.

Febrile episodes and cerebral malaria in and around the time of surgery are not the only concern with malaria. While not well studied, perioperative malaria has been associated with surgical site infections and longer lengths of stay in a small but clinically relevant retrospective study of trauma patients in Southeast Asia. In a review of 227 trauma patients, the rate of wound infection was 36.1% in patients who developed symptomatic malaria within 10 days of surgery compared with 10.0% in patients without symptomatic malaria. In addition, the average hospital stay in the postinjury malaria group was 31.2 days compared with 19.4 days in the patients without the complication.5

It is possible that some parasite protein antigens with prolonged half-lives may have remained positive, and the perioperative fevers in the first group may be from other causes, such as respiratory infections or other common and not-so-common tropical infections.4,6 The manufacturer notes fever as a side effect after administration of artemisinin-based combination therapy.7 Yet, no patient in the second group developed fever during their stay and recovery from surgery, suggesting the possibilities of respiratory illness or drug fever are unlikely.

Another significant problem in many areas of the world is the importation of subtherapeutic and counterfeit medications, including antimalaria medicines, well documented in other studies.8,9 We believe that the antimalarial medications acquired from street vendors for our first team were counterfeit, despite apparently valid labeling from the pharmaceutical manufacturer. However, relapsing malaria may also account for the febrile episodes in the first group of patients, but without an assay, we are not able to firmly conclude that the first batch of medications given to our patients was a forgery.

No studies have been conducted to determine the optimum interval for malaria testing and treatment before elective surgery. The incubation period for the various forms of malaria is 7–28 days. Certainly, 3–4 weeks before elective surgery in group A may allow development of a new episode or a relapse of malaria, particularly in this nutritionally deprived population. On the other hand, our choice of providing universal preventive treatment of malaria, due to the high prevalence in our population, within 3–7 days before elective surgeries, while controversial, is not entirely unique. Chloroquine has been utilized for malaria prophylaxis, rather than treatment, before surgeries, although this medication has fallen out of favor due to high levels of resistance worldwide.10 Intermittent preventive treatment for (presumed) malaria in pregnant patients and in children, discussed previously under Methods, are other examples. Our choice for universal malaria treatment was based on logistical and scheduling reasons for short-term visiting surgical teams, as well as the more rapid clearance of malaria parasites with artemisinin-based combination therapy–based therapies compared to traditional quinine-based treatments. We also had concerns about adverse drug reactions between quinine compounds and our anesthetics, including dysrhythmias and hepatotoxicity and neurotoxicities seen in numerous patients after intramuscular injections. We selected artemisinin-based combination therapy rather than the traditional quinine due to ease of administration (oral versus parenteral), more rapid clearance of parasites, and fewer side effects and toxicity with artemisinin-based combination therapies, as well as the much cheaper cost of administration of oral versus parenteral. One could reasonably posit that, in areas of high malaria transmission, presumptive treatment of malaria with artemisinin-based combination therapies in pediatric patients preoperatively is consistent with and extends the World Health Organization multipronged approaches to malaria eradication (eg, intermittent preventive treatment of [presumed] malaria for pregnant women and intermittent preventive treatment for [presumed] malaria in children) to a greater percentage of adversely affected populations in high-risk areas such as Niger. Until more research is done to determine the optimum timing of antimalarial medications, the World Health Organization still recommends treating only those patients with confirmed malaria diagnosis. However, it seems reasonable to provide treatment to those with a positive malaria test on preoperative testing, possibly a week before elective surgery, to allow some clearance of the medications with longer half lives, but not so long as to allow new infection with malaria. It is unknown if intermittent preventive treatment for (presumed) malaria in children around the time of surgery is effective and can be extended to children >5 years of age, although this retrospective review with small numbers may only suggest this as a possibility. Much more research is needed on these topics, as well as how to best address cryptogenic forms of malaria in the perioperative time frame.

In summary, this study highlights several issues in managing pediatric surgery patients at risk of malaria in the developing world. We report several unexpected findings with our retrospective comparison of these 2 groups of pediatric surgery patients. What is significant about this study was (1) the high overall occurrence of malaria in this pediatric surgical population, particularly nonfalciparum species, despite recent treatment with artemisinin-based therapy; (2) the challenges with limited diagnostic studies for malaria in austere environments; (3) the problems with possible counterfeit medications in the developing world; (4) the timing of administration of antimalaria medications before surgery; and (5) risks attendant with perioperative malaria and its treatment in any surgical patient population. More studies are needed to determine optimum strategies for managing perioperative malaria, particularly with the wide spectrum of malaria presentations throughout the world. One must weigh the risks and benefits of treating all patients or only those with confirmed malaria in regions with high transmission rates of malaria versus the risks of developing drug resistance, the toxicities and drug interactions between antimalaria medicines and various anesthetic agents, and the risks of untreated malaria, including falciparum and cryptogenic malarias in the perioperative time period, particularly in regions of the world where the burden of disease due to surgically treatable conditions is vastly undertreated. These questions remain unanswered.


Name: Gary L. Roark, MD.

Contribution: This author was responsible for providing anesthesia and perioperative services for all the patients described in this retrospective review, as the medical director and as chief of anesthesiology. He received training in anesthesiology and in tropical medicine. The author reviewed all charts, collected the data, and provided the statistical analysis. He was the sole author of this manuscript.

This manuscript was handled by: Angela Enright, MB, FRCPC.


1. Amanor-Boadu SD, Mohammed A. The diagnostic dilemma of intraoperative hyperpyrexia in a malaria endemic area. West Afr J Med. 2003;22:98–100.
2. World Health Organization. World Malaria Report 2018. 2018.Geneva, Switzerland: World Health Organization.
3. Doudou MH, Mahamadou A, Ouba I, et al. A refined estimate of the malaria burden in Niger. Malar J. 2012;11:89.
4. World Health Organization. WHO Treatment Guidelines. Guidelines for the Treatment of Malaria. 2015.3rd ed. Geneva, Switzerland: WHO Press.
5. Sundet M, Heger T, Husum H. Post-injury malaria: a risk factor for wound infection and protracted recovery. Trop Med Int Health. 2004;9:238–242.
6. Abba K, Deeks JJ, Olliaro P, et al. Rapid diagnostic tests for diagnosing uncomplicated P. falciparum malaria in endemic countries. Cochrane Database Syst Rev. 2011;(7):CD008122.
7. (database online). Artemether and Lumefantrine. C1996-2018. Updated November 15, 2018. Available at: Accessed December 21, 2018.
8. Karunamoorthi K. The counterfeit anti-malarial is a crime against humanity: a systematic review of the scientific evidence. Malar J. 2014;13:209.
9. Counterfeit drugs: a growing global threat. Lancet. 2012;379:685.
10. Djibo A, Madougou B, Bougarel J, Chippaux JP. [Value of malaria prophylaxis in surgical intervention in a malaria endemic zone, Niamey, Niger]. Bull Soc Pathol Exot. 2001;94:258–259.
Copyright © 2019 International Anesthesia Research Society