Measuring insecticide resistance in a vacuum: exploring next steps to link resistance data with mosquito control efficacy

Abstract

Insecticide resistance is a great challenge facing mosquito operational control agencies across the United States, where few active ingredients with unique modes of action are available for use, increasing resistance pressure and further hampering resistance management strategies. Emergence and expansion of insecticide resistance in mosquitoes can be detected by resistance monitoring programs; however, there are gaps in our knowledge regarding the link between resistance bioassay results and operational control outcomes. Here, we review both public health and agricultural studies on pesticide resistance bioassays and control outcomes. A discussion on the main gaps in our knowledge of insecticide resistance and a review of resistance management practices is also presented. We conclude with research questions that can advance our understanding of resistance monitoring and control.

Introduction

Pesticide resistance monitoring is an important component of vector management. Since the National Association of County and City Health Official’s (NACCHO) publication of a report showing 98% of vector control programs were lacking in the core competency of pesticide resistance testing, there has been an increase in awareness and efforts to increase this capacity have been significant (NACCHO 2017). Several programs, including the Northeast Regional Center for Excellence in Vector-Borne Diseases (NEVBD) Insecticide Resistance Monitoring Program, have documented resistant and susceptible mosquito vector populations in recent years (Burtis et al. 2021, NEVBD 2023). However, the link between detected insecticide resistance and operational control outcomes is unclear due to few publications analyzing this link among either public health vectors or agricultural pests.

Globally, insecticide resistance is a major concern due to the presence of resistance in mosquito vector species, including Aedes aegypti (Diptera: Culicidae, Linnaeus) (Vazquez-Prokopec et al. 2017, Gray et al. 2018), the vector of dengue, yellow fever, chikungunya, and Zika (Kotsakiozi et al. 2017), and Anopheles spp. (Diptera: Culicidae), the primary vector of the malaria parasite (N’Guessan et al. 2003, Toé et al. 2014, Kleinschmidt et al. 2018, Oumbouke et al. 2020, Meiwald et al. 2022). In North America, resistance has been noted in 2 important vector species of West Nile virus, Culex quinquefasciatus (Diptera: Culicidae, Say) and Culex pipiens (Diptera: Culicidae, Linnaeus) (Paul et al. 2005, Fuseini et al. 2019, Lee et al. 2023). Given the increase in resistance among mosquitoes, there is growing global concern over the resurgence of mosquito populations and mosquito-borne illness (Ranson and Lissenden 2016, Gould et al. 2018).

Insecticide resistance is also a concern within the agricultural industry, with more than 500 agriculturally relevant insect and mite species resistant to at least one insecticide, making their control increasingly challenging (Gould et al. 2018, Van Leeuwen et al. 2020, Mota-Sanchez and Wise 2023). Resistance presents a significant threat to the agricultural industry, with resistance management costing billions of US dollars per year (Gould et al. 2018). Effective management strategies are necessary to address the increasing concern of resistance among both agricultural pests and public health vectors.

Detection and successful management of insecticide-resistant mosquito populations depend on a comprehensive understanding of how to accurately measure resistance and how to interpret resistance data. The CDC bottle bioassay and the WHO tube assay are standardized phenotypic insecticide resistance bioassays commonly used for testing adult mosquitoes (Namias 2021). However, these assays have limitations, including the lack of accounting for environmental variables that influence control in a field setting, such as temperature, humidity, food availability/quality, preexisting pesticide exposure, or sublethal exposure, nor do they account for variations in age or gonotrophic status (Alout et al. 2017, Namias et al. 2021). These limitations make it challenging to delineate whether bioassay results accurately predict the efficacy of mosquito control interventions in the field (Hemingway et al. 2013, Thomas and Read 2016, Alout et al. 2017, Vontas and Mavridis 2019, Grossman et al. 2020, Namias 2021).

Knowing which strategies to implement to effectively control resistant populations is essential in the prevention of vector-borne diseases, as insecticides remain the primary tool available for control. However, there is limited guidance on insecticide resistance management, especially for vectors of public health significance.

Here, we investigated the link between phenotypic insecticide resistance data and operational control outcomes in mosquitoes. In an effort to understand whether other methodologies could be applied to mosquito resistance management and control, we also reviewed reports on other public health vectors and agricultural pests.

In this review, we (i) present results of a literature review on the link between bioassay resistance data and operational control outcomes among mosquitoes, other public health vectors, and agricultural pests; (ii) identify knowledge gaps regarding the interpretation and measurement of insecticide resistance; (iii) discuss the limitations of current resistance monitoring strategies; and (iv) propose research agenda items to address these knowledge gaps and limitations.

Materials and Methods

We conducted a literature review using PubMed and Web of Science. We found few studies that have directly compared mosquito resistance data with operational control efficacy. We then continued our review with the inclusion of papers on other public health vectors and agricultural pests in addition to mosquitoes. The search term “insecticide resistance control failure” in PubMed yielded 310 articles. Using the search terms (“insecticide resistance” OR “pesticide resistance”) AND (“control failure” or “control efficacy”) in a collective Web of Science search yielded 326 results, of which 69 were duplicates of the PubMed search. These papers were manually scanned to identify additional articles. Through this process, a total of 576 articles were screened for inclusion. Searches were not restricted by year of publication. Articles were imported to Covidence (Veritas Health Innovation, Melbourne, Australia) for inclusion screening. Unpublished reports, master’s theses, and doctoral dissertations were excluded. Articles were included if the following were reported: (i) phenotypic bioassay resistance testing results and (ii) efficacy results of insecticide application in either the field, greenhouse, or laboratory setting. Articles were excluded if bioassays were performed on specimens more than 3 generations removed from field-collected samples, unless they were exposed to insecticide each generation to maintain resistance, or if no generation data was reported. Additionally, studies were excluded if there were only anecdotal reports of control failure, without supporting field, greenhouse, or laboratory trial data, or if the method of insecticide application was not included.

In total, we screened 576 articles, 32 (6%) articles met our inclusion criteria and were reviewed for insecticide resistance and control efficacy data.

Results

Most papers did not meet our inclusion criteria because only resistance data was reported, and operational control efficacy was not assessed via field, greenhouse, or laboratory trials. Characteristics of the papers that met our insecticide resistance and control efficacy criteria are summarized in Table 1. The studies covered years from 1973 to 2023 and were conducted primarily in North America (44%) and Africa (28%). Most papers focused on resistance profiles for agricultural pest species (n = 18). This was followed by mosquitoes (n = 12); 8 of the studies reported results on Anopheles spp. (Diptera: Culicidae), the genus that transmits human malaria parasites (Beier 1998). A slight majority of studies (53%) meeting our search criteria investigated resistance within the agricultural sector as opposed to the public health realm. Pyrethroids were the most frequently tested class for insecticide resistance, followed by organophosphates. The WHO tube test was the most commonly used bioassay method, with 25% of studies reporting the use of this test, followed by the CDC bottle bioassay/modified CDC bottle bioassay (16%) (Table 1) (CDC 2012, WHO 2016a).

Table 1.

Key features of studies reviewed (n = 32)

Characteristic	n (%)
Region
North America	14 (44%)
Africa	9 (28%)
South America	3 (9%)
Oceania	3 (9%)
Europe	2 (6%)
Asia	1 (3%)
Sector
Agriculture	17 (53%)
Public health	15 (47%)
Specimen
Agricultural pest	18 (56%)
Mosquito	12 (38%)
Non-mosquito vector	2 (6%)
Insecticide classa
Pyrethroid	22 (69%)
Organophosphate	9 (28%)
Carbamate	3 (9%)
Neonicotinoid	1 (3%)
Other	8 (25%)
Bioassay methoda
WHO tube test	8 (25%)
CDC bottle bioassay	5 (16%)
Topical assay	4 (13%)
Other	16 (50%)

^aSome papers tested more than one insecticide class and/or used more than one bioassay method.

Next, we present a review of the papers that met our inclusion criteria and include a summary of the insecticide resistance data and control results; studies are grouped by organism, starting with (i) mosquitoes, (ii) other public health vectors, and (iii) agricultural pests. Few of the papers that met our inclusion criteria focused on mosquitoes; thus, we included additional organisms to glean information that may be applicable to mosquito control.

The resistance ratio (RR) is a measure of the level of resistance among a test population and is often calculated when monitoring for resistance. It allows for the comparison of resistance across multiple populations (Ninsin 2017). RR was a commonly reported measure in the papers reviewed here. RR is typically calculated by dividing the median lethal concentration (LC50) or median lethal dose (LD50) of a test or field-collected mosquito strain by the LC50 or LD50 of a susceptible strain (WHO 2016b). In 1927, Trevan first introduced “median lethal dose” which is now a measure commonly used to assess the toxicity of pesticides and a component in determining RR (Trevan 1927, Pillai 2021). In the entomology field, reports of using the median lethal concentration or dose and RR to assess resistance date back to at least the 1950s (Busvine 1951, 1956). RRs are used by the WHO to classify resistance intensity in mosquitoes as follows: RR < 5 indicates susceptibility, RR between 5 and 10 indicates moderate resistance, and RR > 10 indicates high resistance (WHO 2016b). To our knowledge, there is no equivalent RR classification for other vector species or agricultural pests. In this review, we apply the WHO RR classification, when RR is reported for mosquito and nonmosquito arthropod species, as a guide for assessing resistance levels.

Mosquitoes

Anopheles (Diptera: Culicidae)

Eight studies investigated insecticide resistance and control efficacy in Anopheles spp. that transmit the malaria parasite Plasmodium falciparum (Plasmodiidae), with a focus on long-lasting insecticidal nets (LLIN) for controlling resistant Anopheles spp. Four studies conducted either WHO tube tests or CDC bottle bioassays to assess resistance in addition to performing laboratory trials with the WHO cone bioassay to determine the impact of resistance on susceptibility to LLINs (N’Guessan et al. 2003, Toé et al. 2014, Oumbouke et al. 2020, Meiwald et al. 2022). When performing the WHO cone bioassay, the WHO categorizes LLINs as effective if there is ≥80% mortality or 95% knockdown (WHO 2013). One study used adapted CDC bottle bioassays and found high resistance to deltamethrin (pyrethroid) (RR = 1,441–2,405) (Oumbouke et al. 2020). These same Anopheles gambiae s.l. (Diptera: Culicidae, Giles) populations were associated with low mortality (<30%) after exposure to deltamethrin-treated LLINs via the WHO cone bioassay (Oumbouke et al. 2020). Another study performed the WHO tube test on field-collected An. gambiae from Cote d’Ivoire and documented carbamate resistance, noting 29% mortality after performing the bioassay (N’Guessan et al. 2003). After exposure to carbamate-impregnated netting in the laboratory, there was only a 12% mortality rate among the resistant mosquito population (N’Guessan et al. 2003). Additionally, Toé et al. used the CDC bottle bioassay (RR > 1,000) and the WHO tube test (RR = 730) to quantify resistance levels to deltamethrin among field-collected adult An. gambiae in Burkina Faso. They found high pyrethroid resistance led to decreased net efficacy when assessed using the WHO cone bioassay (Toé et al. 2014).

However, results from field-based trials show nets can still have a protective effect (i.e., prevent human cases of malaria) in areas with known populations of resistant mosquitoes. For example, Kleinschmidt (2018) compared malaria epidemiological outcomes across geographic clusters with low to high frequencies of pyrethroid-resistant mosquitoes. They found that the use of pyrethroid LLINs protected against human malaria infection, despite Anopheles’ resistance to pyrethroids, compared to nonnet users. However, this study assessed resistance via the WHO tube test and only measured resistance frequency (defined as the proportion of the population that is resistant), not resistance intensity (defined as the strength of resistance) (Dennehy and Dunley 1993, WHO 2022). Consequently, it is not clear if the study outcome would be similar in a setting with high-intensity pyrethroid resistance. Nonetheless, it appears LLINs may continue to offer some level of protection against malaria infection even if insecticide resistance is present in the local mosquito population (Kleinschmidt et al. 2018).

Aedes aegypti (Diptera: Culicidae, Linnaeus)

Two studies that met our criteria focused on Ae. aegypti, the primary vector of dengue, chikungunya, yellow fever, and Zika viruses (Kotsakiozi et al. 2017). A study in Mexico demonstrated a link between resistance and vector control efficacy (Vazquez-Prokopec et al. 2017). Using CDC bottle bioassays, high levels of resistance to deltamethrin were observed among the local Ae. aegypti population (Vazquez-Prokopec et al. 2017). Indoor residual spraying (IRS) with deltamethrin was conducted to determine how resistance affected control; IRS treatment did not reduce Ae. aegypti populations compared to untreated control houses demonstrating IRS treatment failed to control the local pyrethroid-resistant Ae. aegypti population (Vazquez-Prokopec et al. 2017). Similarly, Gray and colleagues reported high pyrethroid resistance in field-collected strains of Ae. aegypti in Mexico as assessed by the CDC bottle bioassay. Subsequent semi-field trials using 2 commercial aerosolized insecticides, both with pyrethroids as an active ingredient, showed significantly reduced mortality among resistant strains (10%–13%) as compared to a susceptible strain (50.1%) (Gray et al. 2018). Both studies indicate mosquito control failed due to high levels of pyrethroid resistance.

Culex (Diptera: Culicidae)

Two studies focused on Culex species fulfilled our inclusion criteria. One study exclusively examined Culex. quinquefasciatus (Diptera: Culicidae, Say) (Lee et al. 2023), while the other investigated both PCR-confirmed Cx. quinquefasciatus (90%) and unidentified Culex species (10%) (Fuseini et al. 2019). Culex quinquefasciatus is widely distributed and known to spread lymphatic filariasis and numerous arboviruses (Manimegalai and Sukanya 2014); it is the primary vector of West Nile virus in the southern United States (Farajollahi et al. 2011, Lee et al. 2023). In Equatorial Guinea, field-collected Culex spp. were resistant to 4 insecticide classes when tested by the WHO tube test. After IRS treatment with pirimiphos-methyl (organophosphate), mortality of the resistant Culex spp. was 5.4% at 2 months postspray compared to over 80% mortality in a susceptible An. gambiae strain (Fuseini et al. 2019). Additionally, Lee and colleagues calculated RRs of pyrethroid-resistant Cx. quinquefasciatus utilizing a laboratory-coated glass vial bioassay, similar to the CDC bottle bioassay, and a cage field trial (Lee et al. 2023). The RR of the resistant population, as determined in the field trial (RR ≥ 3.51) was similar to the ratio derived from the laboratory vial assay (RR = 3.91) (Lee et al. 2023). These findings suggest RRs obtained through laboratory bioassays can be predictive of those acquired through field evaluations for this species, although more studies like this need to be conducted.

Other Public Health Vectors

Triatoma infestans (Hemiptera: Reduviidae, Klug in Meyen)

Two nonmosquito public health vector papers met our criteria. The studies detected insecticide resistance among Triatoma infestans, the kissing bug that transmits the Chagas disease parasite (Klotz et al. 2014). Both investigations conducted topical assays with deltamethrin on first-instar nymphs, whose parents were collected from the field, and found moderate pyrethroid resistance levels (Gurevitz et al. 2012, Gaspe et al. 2021). Gurevitz and colleagues observed a RR of 7.17 (range 4.47–11.50), indicating moderate resistance. They then conducted IRS with deltamethrin to control T. infestans in homes and noted persistent infestation of adults and 5th-instar nymphs 4-months postspray, suggesting IRS treatment did not successfully control the population (Gurevitz et al. 2012). Gaspe and colleagues performed a similar field trial with IRS of deltamethrin and found more than 25% of treated homes had persistent triatomine infestations after treatment. Failure to suppress moderately resistant wild T. infestans populations in homes treated with IRS was correlated to reduced susceptibility to deltamethrin (Gaspe et al. 2021).

Agricultural Pests

More than half (53%) of the studies that met our inclusion criteria were from the agricultural sector (Table 1). These studies investigated resistance in a range of pest species, from food crop pests to parasitic mites and ticks of agricultural animals (Table 2). Various bioassays were used to assess resistance among agricultural pests in contrast to the 3 standardized bioassays (CDC bottle bioassay, WHO tube test, and the larval packet test) applied to arthropods of public health concern (Table 2). Pyrethroids were the most commonly investigated insecticide class, followed by organophosphates.

Table 2.

Studies investigated connections between insecticide resistance and control efficacy

Study type	Species	Insecticide	Insecticide class	Resistance bioassay	Trial type	Reduced efficacy	References
Public health
	Aedes aegypti	Tetremethrin Allethrin Phenothrin Cypermethrin Imiprothrin	Pyrethroid Pyrethroid Pyrethroid Pyrethroid Pyrethroid	CDC bottle	Semifield	Yes	Gray et al. (2018)
	Aedes aegypti	Deltamethrin	Pyrethroid	CDC bottle	Field	Yes	Vazquez-Prokopec et al. (2017)
	Anopheles gambiae	Deltamethrin	Pyrethroid	WHO tube	Semi-field	Yes	Hughes et al. (2020)
	Anopheles gambiae	Permethrin Pyriproxyfen	Pyrethroid Pyridine	WHO tube	Field	Yes	Koffi et al. (2015)
	Anopheles gambiae s.l.	Deltamethrin Deltamethrin + PBO	Pyrethroid Pyrethroid + synergist	CDC bottle	Lab	Yes	Meiwald et al. (2022)
	Anopheles gambiae	Carbosulfan	Carbamate	WHO tube	Lab	Yes	N’Guessan et al. (2003)
	Anopheles gambiae	Deltamethrin Pirimiphos-methyl	Pyrethroid Organophosphate	WHO tube	Field	Yes	Ngufor et al. (2014)
	Anopheles gambiae	Deltamethrin	Pyrethroid	Modified CDC bottle WHO tube	Lab	Yes	Oumbouke et al. (2020)
	Anopheles gambiae	DDT Deltamethrin Permethrin	Organochlorine Pyrethroid Pyrethroid	CDC bottle WHO tube	Lab	Yes	Toé et al. (2014)
	Anopheles spp.	Permethrin Deltamethrin	Pyrethroid Pyrethroid	WHO tube	Field	No	Kleinschmidt et al. (2018)
	Culex quinquefasciatus	Permethrin + PBO	Pyrethroid + synergist	Other: vial assay	Field	Yes	Lee et al. (2023)
	Culex spp.	Pirimiphos-methyl	Organophosphate	WHO tube	Field	Yes	Fuseini et al. (2019)
	Triatoma infestans	Deltamethrin	Pyrethroid	Topical	Field	Yes	Gaspe et al. (2021)
	Triatoma infestans	Deltamethrin	Pyrethroid	Topical	Field	Yes	Gurevitz et al. (2012)
Agriculture
	Bactrocera oleae	Alpha-cypermethrin	Pyrethroid	Topical	Field	Yes	Kampouraki et al. (2018)
	Boophilus microplus	Permethrin	Pyrethroid	Other: larval packet	Field	Yes	Davey and George (1998)
	Bovicola ovis	Cypermethrin Alpha-cypermethrin Deltamethrin	Pyrethroid Pyrethroid Pyrethroid	Other: in vitro technique	Field	Yes	Johnson et al. (1992)
	Delia antiqua	Chlorpyrifos	Organophosphate	Other: larval immersion and topical spray	Field	Yes	Nault et al. (2006)
	Frankliniella fusca	Lambda-cyhalothrin	Pyrethroid	Other: treated foliage	Field	Yes	Krob et al. (2022)
	Haematobia irritans	Diazinon	Organophosphate	Other: impregnated filter paper	Field	Yes	Barros et al. (2001)
	Halotydeus destructor	Chlorpyrifos Omethoate	Organophosphate Organophosphate	Other: coated vial	Field	Yes	Umina et al. (2023)
	Helicoverpa armigera Spodoptera exigua	Broflanilide Chlorantraniliprole Emamectin benzoate	Meta-diamide and isoxazoline Anthranilic diamide Avermectin	Other: leaf dip	Field	Yes	Tang et al. (2021)
	Leucoptera coffeella	Chlorpyrifos	Organophosphate	Other: treated foliage	Greenhouse	Yes	Amaral Rocha et al. (2022)
	Lygus lineolaris	Dicrotophos Methyl parathion Permethrin	Organophosphate Organophosphate Pyrethroid	Other: coated vial and spray chamber	Field	Yes	Snodgrass and Elzen (1995)
	Musca domestica	Bioresmethrin Permethrin Deltamethrin Pyrethrins	Pyrethroid Pyrethroid Pyrethroid Pyrethroid	Topical	Field	Yes	Farnham (1984)
	Panonychus ulmi	Chlordimeform hydrochloride	Formamidine	Other: leaf dip	Field	Yes	Croft and McGroarty (1973)
	Platynota idaeusalis	Methomyl	Carbamate	Other: foliar spray	Semi-field	Yes	Hull et al. (1997)
	Striacosta albicosta	Bifenthrin	Pyrethroid	Other: impregnated filter paper	Lab	No	Montezano et al. (2019)
	Tetranychus urticae	Propargite Fluvalinat	Organosulfite Pyrethroid	Other: leaf disk spray	Field	Yes	Goodwin et al. (1995)
	Thrips tabaci	Lambda-cyhalothrin	Pyrethroid	Other: TIBSa	Greenhouse	Yes	Shelton et al. (2003)
	Varroa destructor	Amitraz	Formamidine	Other: coated vial	Field	Yes	Rinkevich (2020)
	Varroa jacobsoni	Amitraz Fluvalinate	Formamidine Pyrethroid	Other: coated vial	Field	Yes	Elzen et al. (2000)

^aThrips Insecticide Bioassay System.

In a study on onion thrips, Thrips tabaci (Thysanoptera: Thripidae, Lindeman), a common pest of onions, met our inclusion criteria (Shelton et al. 2003). Susceptibility to the pyrethroid lambda-cyhalothrin was assessed using a novel bioassay known as Thrips Insecticide Bioassay System (TIBS) (Rueda and Shelton 2003, Shelton et al. 2003). To assess onion thrips for resistance, specimens were placed in microcentrifuge tubes coated with the formulated product, and mortality was noted after 24 h. Bioassay results were followed by a greenhouse trial during which pyrethroid-resistant and susceptible onion thrips populations were exposed to lambda-cyhalothrin-treated leaves and assessed for mortality; the same onion thrips population was used for both bioassay and treatment trials. A positive linear relationship was observed between bioassay mortality and mortality during the greenhouse trial, indicating that TIBS could be used to predict the likelihood of operational spray success (Shelton et al. 2003). Delineating if a relationship exists between bioassay mortality and control mortality can indicate if control methods with the tested active ingredient will be successful. Understanding this relationship for mosquito vectors could inform control practices and resistance management strategies.

In a study on the western bean cutworm (Striacosta albicosta) (Lepidoptera: Noctuidae, Smith), eggs and adults were collected from fields reporting control failure after bifenthrin (pyrethroid) aerial application. Larvae hatched from field-collected eggs, and adults were exposed to bifenthrin-treated filter paper. The RRs from the bioassay tests ranged from 8.72 to 11.93; however, when a laboratory-based aerial spray simulation was employed to assess the control efficacy of bifenthrin on these resistant populations of S. albicosta, 100% mortality and no control failure was observed (Montezano et al. 2019). These results indicate resistance levels may not have been high enough to reduce the performance of the formulated product in a controlled laboratory environment or that S. albicosta mortality observed during laboratory-based aerial sprays is not reflective of control in the field (Montezano et al. 2019). This suggests laboratory spray simulation results do not always accurately predict the likelihood of control failure.

In another study, Umina and colleagues (2023) investigated the relationship between laboratory-measured pesticide resistance and its impact on field-based pest control for the redlegged earth mite, Halotydeus destructor (Trombidiformes: Penthaleidae, Tucker). The bioassay data demonstrated different levels of resistance for the 2 chemicals of interest: high chlorpyrifos (organophosphate) resistance (RR = 99.3) and moderate omethoate (organophosphate) resistance (RR = 6.6). Subsequent field trials to evaluate the effectiveness of these pesticides for control were consistent with laboratory findings showing differences between the 2 pesticides. For example, following foliar spraying of chlorpyrifos, only a small decline in the resistant mite population occurred compared to the untreated controls. However, omethoate was highly effective and reduced mite abundance by more than 98% (Umina et al. 2023). Ultimately, chlorpyrifos treatment failed to successfully control the mite population, while omethoate did effectively control H. destructor. These results demonstrate that operational control failure does not necessarily occur when moderate levels of resistance are detected, highlighting the importance of accurately measuring resistance levels in the laboratory to inform and improve field-based pest control strategies for the redlegged earth mite.

In a study on tarnished plant bugs (Lygus lineolaris) (Hemiptera: Miridae, Palisot), field-collected specimens were tested for insecticide resistance with a coated glass vial bioassay (Snodgrass and Elzen 1995). Bioassay results demonstrated susceptibility to dicrotophos (organophosphate) and moderate resistance to methyl parathion (organophosphate) with RRs of 2.3 and 8.4, respectively. A laboratory cage trial was conducted under optimal conditions where adult L. lineolaris was exposed to spray-treated cotton. Mortality among adults exposed to dicrotophos and methyl parathion was 100% and 80.6%, respectively. In a field trial targeting the population previously tested in the lab, dicrotophos and methyl parathion were applied to cotton plants at a rate of 55.9 L/ha at 3308 g/cm². Dicrotophos treatment resulted in a 56.7% reduction of L. lineolaris compared to the untreated control; methyl parathion resulted in a 27.9% reduction (Snodgrass and Elzen 1995). These results indicate control of the susceptible and moderately resistant tarnished plant bugs under optimal laboratory conditions, but they did not align with or predict control success under field conditions, where mortality was lower compared to the laboratory control results (Snodgrass and Elzen 1995).

Another study that met our inclusion criteria investigated the control of Varroa destructor (Mesostigmata: Varroidae, Anderson and Trueman), the varroa mite, an external parasite of honeybees (Rinkevich 2020). Varroa mites collected from multiple field sites were exposed to amitraz, a formamidine, during resistance bioassays, and a range of RRs indicating no resistance to high resistance were identified. In subsequent trials, varroa was exposed to Apivar strips (3.3% amitraz). Operational control failure was more likely to occur in populations with high amitraz resistance (>10-fold RR) as compared to populations with lower RRs. In this study, control efficacy was significantly correlated with insecticide RRs, indicating that RRs could be used to predict operational control outcomes (Rinkevich 2020).

Results from several additional agricultural studies included in this review suggest bioassay data, such as RRs and lethal concentrations (LC) can provide insight into operational control outcomes (Farnham 1984, Johnson et al. 1992, Hull et al. 1997, Kampouraki et al. 2018, Tang et al. 2021, Amaral Rocha et al. 2022).

Two studies noted a relationship between lethal concentration and control failure. For example, a study on the coffee leaf miner (Leucoptera coffeella) (Lepidoptera: Lyonetiidae, Guerin-Meneville), a common pest of the coffee plant, evaluated chlorpyrifos resistance and control failure (Amaral Rocha et al. 2022). A modified bioassay was used to expose adult coffee leaf miners to spray-treated coffee seedlings (Gonring et al. 2019). Control failure occurred when the LC₈₀ was greater than the recommended field application rate (Amaral Rocha et al. 2022). Additionally, Hull and colleagues (1997) observed an inverse relationship between LC₅₀ and percent field mortality among larval tufted apple bud moths (Platynota idaeusalis) (Lepidoptera: Tortricidae, Walker) treated with methomyl (carbamate).

Additionally, several studies used bioassays to determine RRs and reported corresponding operational control outcomes (Farnham 1984, Johnson et al. 1992, Goodwin et al. 1995, Kampouraki et al. 2018). One study on the house fly, Musca domestica (Diptera: Muscidae, Linnaeus), demonstrated field-collected flies tested for insecticide resistance via topical assays were not successfully controlled in the field when RRs were at the following levels: bioresmethrin (pyrethroid) RR = 20, permethrin RR = 15, deltamethrin RR = 30, and pyrethrins RR = 4 (Farnham 1984). Additionally, Kampouraki and colleagues (2018) investigated insecticide resistance among olive fruit flies (Bactrocera oleae, Diptera: Tephritidae, Rossi). In field cage trials, populations were exposed to the pyrethroid alpha-cypermethrin (pyrethroid) via a 0.3 L water solution containing 300 mg L⁻¹ of alpha-cypermethrin, the diagnostic dose. The mortality of the highly resistant strain (RR = 331) was 1.6% on day 1 and 24% on day 4 compared to 10.1% and 43.4% among the more susceptible strain (RR = 36) (Kampouraki et al. 2018). In another study, the susceptibility of the twospotted spider mite, Tetranychus urticae (Trombidiformes: Tetranychidae, Koch), to fluvalinate (240 g/L EC) (pyrethroid) and propargite (570 g/L EC) (organosulfite) was assessed by exposing field samples to treated bean leaves (Goodwin et al. 1995). In a field trial, roses were sprayed with commercially recommended rates of fluvalinate (40 ml/100 L) and propargite (100 g/100 L). Control failure occurred among highly resistant populations treated with fluvalinate and propargite with RRs of 23–51 and 88–135, respectively (Goodwin et al. 1995). Finally, a pour-on pyrethroid treatment applied to sheep infested with pyrethroid-resistant biting lice (Bovicola ovis, Phthiraptera: Bovicoliidae, Schrank) failed to eliminate the infestation when the RR was greater than 4, indicating a correlation between RR and operational control efficacy (Johnson et al. 1992). Collectively, these studies demonstrate multiple instances where bioassay outcomes such as RR and LC were associated with control outcomes and likelihood of control success.

Overall, our search revealed a small number of mosquito studies that noted high pyrethroid resistance intensity in An. gambiae (Toé et al. 2014, Oumbouke et al. 2020) and Ae. aegypti (Vazquez-Prokopec et al. 2017, Gray et al. 2018) resulted in poor vector control in the field. Four agricultural studies also noted high resistance levels in various pest species, which resulted in control failure (Goodwin et al. 1995, Kampouraki et al. 2018, Rinkevich 2020, Umina et al. 2023). Additionally, 2 papers on triatomine bugs showed moderate pyrethroid resistance resulting in decreased mortality in field trials (Gurevitz et al. 2012, Gaspe et al. 2021). We also reviewed an agricultural pest study that demonstrated moderate levels of organophosphate resistance did not negatively impact operational control (Umina et al. 2023). Similarly, a study on the tomato pinworm (Tuta absoluta, Lepidoptera: Gelechiidae, Meyrick), which did not meet our inclusion criteria due to the use of specimens for bioassay testing that were reared in the laboratory for greater than 3 generations, showed low levels of resistance to abamectin (avermectin) and spinosad (spinosyn) did not result in control failure (Silva et al. 2011). However, it also showed low-level resistance to multiple pyrethroids, which resulted in field control failure (Silva et al. 2011). These outcomes suggest high to moderate levels of insecticide resistance may result in control failure, while it may be difficult to predict if control failure will occur if low-level resistance is detected.

Discussion

The goal of this literature review was to identify peer-reviewed studies that delineated a link between insecticide resistance bioassay data and operational control outcomes. Our findings show only a small proportion of published studies have addressed this link. Additionally, resources, such as reference books, on the topic of insecticide resistance address resistance detection, mechanisms, and management but not the link of interest here (Brown et al. 1971, Roush and Tabashnik 1990, Onstad 2008). While the results of our literature review indicate certain bioassay metrics (e.g., RR, LC) may be useful in predicting control efficacy, this relationship is not well defined, particularly for human disease vectors.

Overall, of the 576 articles screened in this review, only 32 met our inclusion criteria and addressed operational control outcomes after the detection of pesticide resistance (Table 2). We recognize knowledge gaps remain regarding the interpretation and measurement of insecticide resistance as well as linking resistance to operational control efficacy (presented in Fig. 1).

Fig. 1.

Gaps in our understanding of the linkages between resistance test results and control efficacy.

Knowledge Gaps

We noted several knowledge gaps regarding the interpretation and measurement of insecticide resistance. These gaps can be grouped into 4 topical areas: resistance bioassays, population resistance dynamics, reference strains, and spatial and temporal factors (Fig. 1). We discuss the relevant gaps in more detail below.

Correlation between bioassay results and control outcomes

The CDC bottle bioassay and the WHO tube test are widely used for the detection of phenotypic evidence of insecticide resistance among various mosquito species against a variety of different insecticides. In the studies described here, the WHO tube test (25%) and the CDC bottle bioassay/modified CDC bottle bioassay (16%) were the most frequently used resistance bioassays. There is a need to refine how bioassay data is analyzed to determine whether bioassay resistance data and control outcome (e.g., mortality) are correlated. Only one paper reviewed here formally described a correlation between bioassay results and control outcomes; this was accomplished using onion thrips and a coated microcentrifuge assay (Rueda and Shelton 2003, Shelton et al. 2003).To our knowledge, no similar correlation has been described for mosquitoes using either of the standard bioassays. If addressed, this research gap could reliably inform and improve mosquito control interventions.

Population-level resistance dynamics

Our current understanding of insecticide resistance among mosquito populations is hampered by several additional knowledge gaps. One crucial gap relates to the need to understand how long it may take for a resistant mosquito population to return to susceptibility in the field once chemical exposure ceases. While this has been investigated for Ae. aegypti in a laboratory setting (with increased susceptibility observed after 10–15 generations) (Chang et al. 2012, Brito et al. 2013, Grossman et al. 2018), there is limited evidence for reversion among mosquito field populations (Parker-Crockett et al. 2022). As a resistance management strategy, many control programs are instructed to rotate chemical classes, but rotation will not be effective if mosquitoes retain high resistance levels over a rotation period. Additionally, there is a gap in understanding what mosquito resistance levels correlate with epidemiological outcomes, such as an increase in human disease cases (Toé et al. 2014, Bagi et al. 2015). We note an additional gap in understanding the role modeling can play in deciphering the dynamic between resistance and control.

Issues with susceptible strains

Using laboratory-reared reference strains of mosquitoes is widely regarded as the gold standard for insecticide susceptibility testing and allows for standardization among tests. However, these strains are often highly inbred, rendering them less representative of the actual baseline susceptibility present in local field populations, yet the existence of entirely unexposed field strains for use in susceptibility testing is unlikely. Nonetheless, the ultimate goal is to assess whether resistance is evolving over time and if it impacts effective mosquito control measures. To establish a reliable baseline for comparison, it is advisable to have a diverse range of susceptibility data to gauge the status of wild populations accurately.

Need for a better understanding of sampling scope and frequency

It is imperative to determine appropriate methods for conducting resistance testing in order to accurately assess the geographic scale of mosquito susceptibility in a population targeted for control. Additionally, questions remain regarding the optimal sampling frequency for resistance testing. Answering these questions will inform the implementation of efficient and successful resistance monitoring programs. Historically, few vector control programs in the United States performed pesticide resistance testing (NACCHO 2017, Burtis et al. 2021). Initiating a resistance monitoring program is costly, time consuming, and requires trained personnel. Answering the questions of sampling scope and resistance testing frequency could reduce barriers to initiating successful resistance monitoring programs.

Limited Resistance Management Guidance

We also note the limited resources available for decision-making once a resistant mosquito population is detected. Guidance on best practices for performing resistance monitoring and conducting resistance bioassays (IRAC 2011, NEVBD 2023, WHO 2022) is available, but information is scant on how to proceed once resistance is detected in public health vectors.

The topic of insecticide resistance, including resistance monitoring, detection, mechanisms, and management, has been well described in the agricultural sector (Roush and Tabashnik 1990, Onstad 2008). Specifically, resistance management includes chemical strategies such as rotating chemical classes, which entails cycling between multiple pesticides with different modes of action (MOA), using chemical mixtures, using less persistent chemicals, changing the frequency of application, and targeting a specific life stage (Roush and Tabashnik 1990, Onstad 2008). Nonchemical approaches include biological controls and landscape changes, such as establishing an untreated refuge for the maintenance of susceptible populations (Roush and Tabashnik 1990, Onstad 2008, Gould et al. 2018). However, some of these approaches, such as establishing a refugium to maintain susceptibility, may not be feasible for public health vectors as they could increase disease transmission risks to humans in the surrounding areas.

The Insecticide Resistance Action Committee (IRAC) provides limited resistance management guidance specifically for vectors of public health concern (IRAC 2011). IRAC recommends some strategies gleaned from agriculture, including rotating chemical classes to prevent or slow the emergence of resistance. This strategy is useful in the agricultural setting where a greater variety of MOAs are available. However, few chemical classes with discrete MOAs are available for public health use in the United States; thus, rotating MOAs is often not a practical option. For example, in many US states, only pyrethroids/pyrethrins are registered for adult mosquito control or, occasionally, organophosphates (EPA 2023). Larval mosquito control can involve a slightly larger, but still limited, range of MOAs, including insect growth regulators and microbial toxins. In addition, the MOAs used for public health vector control are often used in agriculture, leading to potential environmental exposure of vectors to these products (IRAC 2011). Thus, rotating MOAs as a resistance management strategy is often not possible.

IRAC also recommends chemical mixtures, or the simultaneous application of 2 or more insecticides, another strategy commonly used in agriculture. However, this is subject to the same shortcomings described above. IRAC advises on the addition of a synergist if metabolic resistance is present; this approach can increase pesticide efficacy and prolong the use of a chemical even where resistance is present. However, IRAC states that resistance management programs “are most effective if implemented before resistance develops or when resistance gene frequency is still very low” further demonstrating a lack of guidance for tackling resistance when frequency is high (IRAC 2011).

Proposed Research Agenda

Improving resistance monitoring, resistance management, and vector control in the future necessitates a comprehensive approach that addresses several knowledge gaps (Fig. 2). First, there is a need to delineate if current mosquito insecticide resistance bioassay data, such as the CDC bottle bioassay and the WHO tube test, are predictive of control efficacy. Investigating and defining a predictive correlation between bioassay results and field control outcomes would be valuable for the prevention of vector-borne diseases.

Fig. 2.

Future research objectives that could address the knowledge gaps linking resistance results to operational efficacy.

Field data needs to be generated on effective resistance management strategies, and the development of practical recommendations for vector control professionals is essential. Additionally, understanding the temporal frequency and spatial scales for monitoring resistance is imperative to enhance our ability to track insecticide resistance dynamics accurately and to inform mosaics and rotational class approaches across time.

Addressing the challenge of controlling resistant mosquito populations requires the development of rigorous and repeatable methodologies and protocols. To this end, integrating multiple methods, such as laboratory and field-caged trials, could fill some of these knowledge gaps. Moreover, landscape-level experiments, expanding beyond traditional field trials, would be useful to fully understand resistance dynamics (Gould et al. 2018). Further work is also needed to determine whether rotating chemicals leads to a reversion back to susceptibility and how long it takes for reversion to occur for each insecticide class and each mosquito species/strain.

Understanding the impact of insecticide degradation and dilution on resistance emergence is essential to inform evidence-based management strategies. Drawing insights from the agricultural sector’s experience where pesticide dosage variation due to product degradation is observed (Gould et al. 2018), analogous considerations could be applied to IRS in malaria control and larval mosquito control. Additionally, a metric used in agriculture for pest management decision-making is economic-injury level (EIL), which represents “the lowest population density of a pest that will cause economic damage” (Stern et al. 1959). EIL is used as a threshold to determine when pesticides should be used on crops. Similar clearly defined thresholds would be useful in mosquito control for informing management decisions; however, unlike the EIL, which tolerates some level of crop damage before significant economic harm occurs, the stakes for protecting human health are distinct, with even a minimal risk to human well-being deemed unacceptable.

Ultimately, addressing these knowledge gaps could lead to a much-needed comprehensive and evidence-based framework for guiding effective control interventions once resistance is detected. Addressing these knowledge gaps in an interdisciplinary manner could ensure the long-term success of resistance management strategies essential for successfully combatting vector-borne diseases.

Conclusion

Limited reliable evidence linking standard resistance bioassay data to control outcomes in the field limits our ability to effectively control mosquito populations. We reviewed studies that demonstrated reduced control efficacy when high to moderate levels of resistance existed. However, the results were mixed when resistance levels were low. There is a need to identify thresholds at which resistance is likely to result in reduced control efficacy and/or greater epidemiological outcomes for human vector-borne disease. Standardization of methods to quantify resistance would enhance control program decision-making regarding resistance management strategies. Resistance is challenging and multifaceted, making it one of the greatest issues faced by operational mosquito control in the United States and internationally. When resistance is detected, decisions often need to encompass more than changing chemicals/MOAs or investigating genetic factors at play but include economic, political, and social concerns too, creating a complex issue (Gould et al. 2018). Ultimately, new coordinated approaches linking resistance with control efficacy, carefully developed recommendations for resistance management, and identification of novel active ingredients with new MOAs will be essential for protecting the public from vector-borne diseases.

Funding

This work was supported by the Centers for Disease Control and Prevention (U01CK000509 and U50CK2023006600). This content is solely the responsibility of the authors and does not necessarily represent the official views of the Centers for Disease Control and Prevention.

Author Contributions

Áine Lehane (Conceptualization [equal], Formal analysis [lead], Investigation [equal], Methodology [equal], Project administration [lead], Supervision [equal], Visualization [lead], Writing—original draft [lead], Writing—review & editing [equal]), Casey Parker‑Crockett (Writing—review & editing [equal]), Edmund J. Norris (Writing—review & editing [equal]), Sarah Wheeler (Writing—review & editing [equal]), and Laura Harrington (Conceptualization [equal], Funding acquisition [lead], Investigation [equal], Methodology [equal], Resources [lead], Supervision [equal], Visualization [supporting], Writing—review & editing [equal])