Abstract
Background
Vulvovaginal candidiasis (VVC) is commonly managed syndromically or investigated using fungal culture. This study compared direct medical costs of 3 diagnostic pathways for acute VVC in Uganda.
Methods
A decision-analytic cost model compared syndromic management, culture-based diagnosis, and microscopy-based diagnosis in a hypothetical cohort of 1000 symptomatic women. The proportion classified by the microscopy-based framework as meeting the model criteria for antifungal treatment (10.1%) was derived from 447 high vaginal swab records at Sebbi Hospital (2022–2025). Laboratory and antifungal costs were obtained from procurement records and Uganda Joint Medical Stores prices. The culture-based pathway incorporated a return-visit consultation fee. One-way sensitivity analyses were conducted, and a societal-perspective scenario used to assess cost differences between pathways.
Results
Total direct costs per 1000 women were US dollar (USD) 10 761 for syndromic management, USD 35 565 for culture-based diagnosis, and USD 14 996 for microscopy-based diagnosis. Culture-based diagnosis was the costliest pathway under every condition examined, exceeding microscopy-based diagnosis by USD 20 568 and syndromic management by USD 24 804. Microscopy-based diagnosis did not generate net savings over syndromic management at base-case public-sector antifungal prices. The cost difference was most sensitive to the unit cost of microscopy and antifungal treatment. Selective fungal culture increased the cost of the microscopy-based pathway in a separate scenario analysis. For the base-case microscopy-based pathway, without selective fungal culture, the break-even antifungal unit price was USD 6.01 per treatment episode.
Conclusions
Under the modeled assumptions, microscopy-based diagnosis becomes less costly than syndromic management when the antifungal treatment price exceeded the base-case break-even threshold.
Keywords: fungal culture, microscopy, syndromic management, vulvovaginal candidiasis
Vulvovaginal candidiasis (VVC) is among the most common genital infections affecting women of reproductive age, with up to 75% experiencing at least one episode during their lifetime and 6%–9% developing recurrent disease [1]. In Uganda, the annual economic burden of recurrent VVC has been estimated at USD 75.696 million, equivalent to 11.22% of national health expenditure, with direct medical costs comprising 71.5% of this total [2]. Recurrent disease may follow inadequately managed acute episodes, while the cost associated with alternative diagnostic and treatment pathways for acute VVC remains insufficiently quantified.
In many low-resource settings, including Uganda, VVC is commonly managed using syndromic approaches, while laboratory-based diagnosis may involve microscopy or fungal culture where these services are available. Syndromic management relies primarily on symptoms and clinical assessment and may result in empirical antifungal treatment without laboratory confirmation [3]. Fungal culture can identify Candida growth and, where species identification is available, provide additional microbiological information. However, detection of Candida does not by itself establish that symptoms are attributable to active infection rather than colonization [4, 5]. Where Candida detection is used to guide treatment, women with Candida colonization may receive antifungal therapy despite the absence of evidence that Candida is responsible for the presenting symptoms, potentially generating avoidable treatment costs and contributing to antifungal selection pressure [6, 7].
Candida albicans, responsible for the majority of VVC in community settings across sub-Saharan Africa [8], produces pseudohyphae and true hyphae that penetrate vaginal epithelial cells, provoking polymorphonuclear cell (PMN) recruitment [9]. Microscopy can provide information on both Candida morphology and the associated inflammatory response [10, 11]. In this study, we incorporated morphology and inflammatory immune response into a predefined microscopy-based classification framework intended to distinguish women more likely to have acute VVC from those with findings more consistent with Candida colonization or without laboratory evidence of acute VVC. In particular, the framework considered the presence of Candida forms together with neutrophils as supportive of acute VVC, while Candida forms without an accompanying inflammatory response were classified as more consistent with colonization. These classifications represent model assumptions based on microscopy findings and were not independently validated against fungal culture or another reference standard in the study cohort. Notably, this classification is less straightforward for non-albicans Candida species. Species such as Candida glabrata may occur predominantly in yeast forms and may produce a less pronounced inflammatory response, including in asymptomatic women [12]. Consequently, absence of prominent pseudohyphal or hyphal forms does not necessarily exclude clinically relevant non-albicans Candida infection. The microscopy framework therefore included a separate classification for women considered likely to have acute non-albicans Candida infection, recognizing the greater uncertainty associated with this group.
Where laboratory-based diagnosis is available, microscopy and fungal culture may be used to provide additional information beyond syndromic assessment. However, where Candida detection alone is used to guide treatment, women with Candida colonization may receive antifungal therapy despite the absence of evidence that Candida is responsible for their presenting symptoms. This provides a potential rationale for diagnostic approaches that incorporate both organism morphology and inflammatory findings when considering treatment decisions. Pooled regional data have reported a Candida culture positivity rate of 33% among women evaluated in relevant studies [8]. Despite the widespread use of syndromic management and the availability of microscopy in some Ugandan laboratories, the economic implications of alternative diagnostic pathways for symptomatic women have not been well characterized. In particular, the potential costs associated with routine fungal culture, additional consultations, and selective-culture strategies have not been evaluated alongside a microscopy-based classification approach. We therefore developed a decision-analytic model to compare the direct medical costs of syndromic management, routine fungal culture, and microscopy-based diagnosis among symptomatic women in Uganda.
METHODS
Study Design and Rationale
We conducted a decision-analytic cost analysis comparing the direct medical costs of 3 diagnostic pathways for acute VVC: syndromic management, culture-based diagnosis, and microscopy-based diagnosis. The model was constructed from 3 sources of empirical data: routine laboratory data from Sebbi Hospital Laboratory (January 2022 to December 2025), laboratory unit costs derived from procurement records at the same facility, and antifungal treatment costs at current facility dispensing prices. Other model parameters were obtained from published literature or specified as model assumptions and varied in sensitivity analyses as appropriate. The analysis adopted a healthcare-system perspective for the base case and reports costs in US dollars (USD) with Uganda Shilling (UGX) equivalents at the prevailing exchange rate. The 3 pathways differ in the number of clinical encounters they require. Culture-based diagnosis involves a return visit for results, while syndromic management and microscopy-based diagnosis are modeled as requiring a single clinical encounter. The additional burden associated with this return visit was captured separately through a societal-perspective scenario incorporating patient transport and time costs, which were reported alongside the base case rather than within it.
Study Setting
The study was conducted at Sebbi Hospital Laboratory, which serves Sebbi Hospital and surrounding health facilities in Nansana Municipality, Uganda. Sebbi Hospital operates within Uganda's tiered public health system and provides diagnostic laboratory services to outpatient and inpatient populations. Since January 2022, the laboratory has implemented microscopy-based diagnosis as routine practice for all high vaginal swab assessments requested for suspected VVC, using a structured 4-category microscopy-based classification framework.
Data Sources
Routine laboratory data were extracted from Streamline, the electronic laboratory information management system used at Sebbi Hospital, for the period January 2022 to December 2025. The dataset comprises 447 consecutive high vaginal swab records from women evaluated for suspected VVC, each with documented clinical information and microscopy findings (Supplementary Data 1). This dataset was derived from microscopy findings only. Fungal culture was not performed as part of this primary 447-record dataset from which the microscopy-based diagnostic category distribution was derived.
Microscopy-based Diagnostic Framework
The microscopy-based diagnostic framework applied in this study integrates clinical assessment with wet mount and Gram-stain microscopy. Wet mount preparations using 10% potassium hydroxide and saline were examined under ×40 objective. Gram staining was performed by sequential application of crystal violet (1 minute), Gram iodine (1 minute), 50% acetone-alcohol decolourizer (10 seconds), and neutral red counterstain (1 minute). The Gram-stained slides were examined under ×100 oil immersion objective. The presence of blastospores, pseudohyphae or hyphae, and PMNs (predominantly neutrophils) was recorded dichotomously as present or absent.
For the purposes of the decision model, observed microscopy findings were assigned to 4 prespecified model-based classification categories with corresponding treatment implications, as follows:
Likely acute VVC: blastospores present with pseudohyphae or hyphae and neutrophils present. For purposes of the model, this combination was interpreted as consistent with hyphal-associated acute VVC, most commonly associated with C. albicans, and was assigned antifungal treatment. This interpretation was not independently validated against a reference standard in the study dataset.
Likely acute VVC due to non-albicans Candida species: blastospores with neutrophils present, but no pseudohyphae or hyphae. For purposes of the model, this pattern was considered compatible with possible non-albicans Candida infection, including C. glabrata, recognizing that species identity and clinical infection status could not be confirmed in the study dataset. Antifungal treatment was assigned in the model, with selective fungal culture considered where additional species-level or susceptibility information was required.
Likely Candida species colonization: blastospores present without neutrophils. For the purpose of the classification framework, the absence of documented neutrophilic inflammation was interpreted as more consistent with Candida colonization than with acute VVC. Antifungal treatment was not assigned in the model. In the selective-culture scenario, fungal culture for species identification and antifungal susceptibility testing was applied selectively when symptoms persisted despite withholding treatment, to investigate possible non-albicans Candida infection.
No laboratory evidence of acute VVC: no Candida species cells detected. The model assigned no antifungal treatment, while alternative causes of symptoms would require clinical assessment outside the scope of this classification framework.
These 4 categories were model-based classifications derived from observed microscopy findings and prespecified analytical assumptions. Because fungal culture, molecular testing, and another independent reference standard were not available for the 447-record dataset, the framework could not establish diagnostic accuracy or confirm infection or colonization status. Misclassification of women between categories could alter the estimated proportion assigned antifungal treatment and, consequently, the modeled treatment costs and incremental cost differences between pathways.
Despite this, the classification framework was informed by published descriptions of the biological distinction between hyphal-associated invasion and Candida colonization [9, 12]. Notably, interobserver reliability was not assessed for this 4-category framework. The framework is summarized in Figure 1 alongside the cost-consequence model structure.
Figure 1.
Microscopy-based diagnostic pathway and cost-consequence model structure for acute vulvovaginal candidiasis. The upper section shows the clinical diagnostic pathway from patient presentation through wet mount and Gram-stain microscopy, interpretation of microscopy findings, and treatment decision. For the purposes of the model, women are assigned to microscopy-based classification categories according to the recorded presence of Candida elements and neutrophils, with antifungal treatment assigned to those meeting microscopy criteria for acute VVC. Selective fungal culture is reserved for likely non-albicans Candida cases and for women with persistent symptoms following withholding of treatment. The lower section shows the decision-analytic cost-consequence model comparing direct medical costs of syndromic management, culture-based diagnosis, and microscopy-based diagnosis for a hypothetical cohort of 1000 symptomatic women.
Decision-Analytic Model Structure
A static decision-analytic model was constructed to estimate and compare the direct medical costs of 3 diagnostic pathways applied to a hypothetical cohort of 1000 symptomatic women presenting for laboratory evaluation of suspected VVC. The model structure and patient flow are presented in Figure 1. Each pathway comprises 2 cost components: laboratory costs and antifungal treatment costs. Direct costs comprised laboratory costs, antifungal treatment costs, and consultation fees, as applicable to each pathway.
Syndromic Management
Syndromic management represents a modeled scenario in which symptomatic women receive empirical antifungal treatment without laboratory investigation. No laboratory cost is incurred. All 1000 women receive antifungal treatment regardless of infection status.
Culture-based Diagnosis
Culture-based diagnosis models the scenario in which microscopy and fungal culture are both performed for all symptomatic women, and treatment is assumed to follow culture results alone. In the model, all culture-positive women were assumed to receive antifungal treatment regardless of whether microscopy indicates active infection or colonization. Laboratory costs are incurred for the combined culture service, which includes microscopy as part of the routine workup, for all 1000 women at USD 16.22 (UGX 60 000) per test. The proportion of symptomatic women assumed to be culture-positive was 33%, based on pooled regional estimates reported in studies using culture-based diagnosis [8]. Because fungal culture results are not always available at the point of care, women undergoing culture-based diagnosis are assumed to return for a second consultation to receive results and, where indicated, a prescription. This return-visit consultation fee is incurred by all women in this pathway regardless of culture outcome, consistent with routine clinical practice in which patients receive both their positive/negative results from their managing doctors/obstetricians.
Microscopy-based Diagnosis
Microscopy-based diagnosis represents the scenario in which treatment decisions are guided by predefined microscopy classification framework, incorporating the presence of Candida elements and accompanying host inflammatory response (presence of PMNs). Microscopy costs were applied to all 1000 women.
Only women classified as likely acute VVC under the microscopy framework were assigned antifungal treatment in the model, representing 10.1% (95% confidence interval [CI] 7.4%–13.2%) of the distribution derived from the 447-record study dataset. The base-case model assigned no antifungal treatment to women classified as likely Candida colonization or as having no laboratory evidence of acute VVC. Selective culture and potential retreatment were examined separately in scenario analysis.
Incremental Cost
The incremental cost of microscopy-based diagnosis relative to syndromic management was defined as the difference in total direct cost between the microscopy-based diagnosis pathway and the syndromic management pathway, after accounting for the additional laboratory cost of microscopy and the reduction in model-assigned antifungal treatment among women classified as likely acute VVC. A positive value indicates that microscopy-based diagnosis is more costly than syndromic management, while a negative value indicates net cost savings.
Cost Inputs
Direct medical costs comprised laboratory, consultation, and antifungal treatment costs (Table 1). Laboratory unit costs represent local market service costs for the relevant laboratory procedures. Antifungal treatment unit costs were based on Joint Medical Stores procurement prices, with a 60% upward adjustment as previously described by the Medicines Transparency Alliance [15]. To isolate sensitivity to the facility markup itself, rather than the antifungal price, this markup assumption was additionally tested at 30% and 90% in sensitivity analysis. The antifungal cost applied was the arithmetic mean of unit costs of fluconazole and clotrimazole, assuming equal prescribing frequencies of the two agents. Patient-borne costs and productivity losses were not captured in the laboratory dataset and were excluded from the healthcare-system base-case analysis. A scenario analysis incorporating these costs for the culture-based pathway's return visit was reported separately. Thus, the base-case healthcare-system analysis does not capture patient-borne transport and productivity costs across all pathways. All costs were converted at an exchange rate of UGX 3700 per USD.
Table 1.
Model inputs
| Parameter | Base Case (UGX) | Base Case (USD) | Low UGX (USD) | High UGX (USD) | Source |
|---|---|---|---|---|---|
| Laboratory unit costs | |||||
| Microscopy per test (wet mount + Gram stain) | 20 000 | 5.41 | 10 000 (2.70) | 30 000 (8.11) | a |
| Fungal culture per test | 60 000 | 16.22 | 30 000 (8.11) | 90 000 (24.32) | a |
| Antifungal treatment unit costs | |||||
| Fluconazole 200 mg-JMS procurement price per 200 mg capsule | 236 | 0.06 | … | … | b |
| Fluconazole 200 mg-facility dispensing price (60% uplift) | 378 | 0.10 | … | … | Derived |
| Clotrimazole 500 mg pessary-JMS procurement price | 5782 | 1.56 | … | … | b |
| Clotrimazole 500 mg pessary-facility dispensing price (60% uplift) | 9251 | 2.50 | … | … | Derived |
| Weighted mean antifungal cost per episode (50:50 assumed mix) | 4815 | 1.30 | 2407 (0.65) | 24 075 (6.51) | Derived. High = private pharmacy estimate |
| Consultation fees | |||||
| General doctor consultation | 20 000 | 5.41 | … | … | a |
| Obstetrician/gynecologist consultation | 50 000 | 13.51 | … | … | a |
| Weighted mean consultation fee (50:50 assumed split) | 35 000 | 9.46 | 20 000 (5.41) | 50 000 (13.51) | Derived |
| Consultations per woman-syndromic/microscopy-based diagnosis | 1 | … | … | … | Model structure |
| Consultations per woman-culture-based pathway | 2 | … | … | … | Return visit for results |
| Societal-perspective inputs (culture pathway return visit only) | |||||
| Patient transport cost per return visit | … | 1.44 | 0.81 | 2.18 | Shiri et al [13] |
| Patient time cost per return visit (full day lost) | … | 1.46 | 1.46 | 2.08 | UBOS National Labour Force Survey 2021 [14] |
| Selective culture scenario (microscopy-based diagnosis) | |||||
| Microscopy-classified acute VVC cases classified as non-albicans Candida | 13.3% | … | … | … | Study cohort (6/45) |
| Assumed proportion requiring culture after persistent symptoms | 10% | … | 5% | 20% | Model assumption informed by reported recurrence estimates of 6%–9% [1] |
| Epidemiological model parameters | |||||
| Proportion assigned antifungal treatment under microscopy framework | 10.1% | … | 7.4% | 13.2% | Study cohort (Supplementary Data 1), binomial exact CI |
| Assumed culture positivity rate | 33.0% | … | 28.0% | 38.0% | Mushi et al 2022 [8], pooled 95% CI |
| Overtreatment fraction-culture-based pathway | 45.6% | … | 35.9% | 55.7% | Supplementary Data 2 |
| Cohort size | 1000 | … | … | … | Model base unit |
| Exchange rate | 3700 UGX/USD | … | … | … | … |
Abbreviations: CI, confidence interval; JMS, Joint Medical Stores; UGX, Uganda Shilling; USD, US dollar.
aLocal market service costs.
bJoint Medical Stores price list.
Base-Case Healthcare Perspective Model Inputs
The microscopy unit cost was UGX 20 000 (USD 5.41) per test, representing the local market service cost for combined wet mount and Gram-stain microscopy (Table 1). The fungal culture unit cost was UGX 60 000 (USD 16.22) per test. Antifungal treatment costs were derived from Joint Medical Stores procurement prices with a 60% upward adjustment to reflect facility dispensing prices [15]. A fluconazole 200 mg capsule was costed at UGX 378 (USD 0.10) and UGX 9251 (USD 2.50) per clotrimazole 500 mg pessary. The weighted mean antifungal cost per treated episode, assuming equal prescribing frequency of fluconazole and clotrimazole, was UGX 4815 (USD 1.30). The weighted mean consultation fee, assuming equal distribution between general doctor (UGX 20 000) and obstetrician (UGX 50 000) consultations, was UGX 35 000 (USD 9.46). In the culture-based diagnosis pathway, this consultation fee was assumed to be incurred twice per woman: once at the initial visit for sample collection and once at the return visit for results and, where applicable, antifungal prescription. All costs were converted at an exchange rate of UGX 3700 per USD.
Societal-Perspective Scenario Inputs
The societal-perspective scenario added two patient-borne cost components to the culture-based pathway's return visit. Transport cost per visit (USD 1.44, applied as the base case with 95% CI 0.81–2.18 in sensitivity analysis) was derived from an ambulatory cohort study across primary health facilities in Tanzania and Uganda [13]. The value of patient time lost to a full day of clinic attendance was estimated using the human-capital approach, based on the median monthly cash wage for female workers in Uganda (UGX 140 000; Uganda National Labour Force Survey 2021 [14]), divided by an assumed 26 working days per month. The overall median wage (UGX 200 000/month) was used as an upper bound in sensitivity analysis. A full day, rather than a partial day, was assumed to be lost per return visit, reflecting the assumed impact of facility queuing and attendance time.
Model Assumptions
The model rests on the following assumptions, each of which is examined in sensitivity analysis:
All symptomatic women in syndromic management pathway were assumed to receive antifungal treatment. This assumption reflects syndromic management as practiced in many low-resourced settings, where laboratory confirmation is not routinely available before treatment initiation.
All culture-positive women were assumed to receive antifungal treatment regardless of colonization status, reflecting the model structure in which culture positivity alone triggers treatment.
The assumed culture positivity rate among symptomatic women was 33%, based on a meta-analysis across sub-Saharan Africa [8]. This parameter was varied across its reported 95% CI (28%–38%) in sensitivity analysis.
The overtreatment fraction in the culture-based diagnosis was defined in the model as the proportion of culture-positive women without microscopy evidence of likely acute VVC under the microscopy framework. This estimate was derived from a secondary analysis of data from a previous study and was 45.6% (95% CI 35.9%–55.7%). This parameter is varied between 35.9% and 55.7% in sensitivity analysis. It was used to estimate overtreatment in the culture-based pathway and in the separate retreatment scenario. It did not affect the base-case total cost comparison because retreatment costs were excluded from the base case.
Fungal culture costs in the microscopy-based pathway, incurred selectively for women classified as likely non-albicans Candida infection and for women with persistent symptoms after a colonization classification, were examined in a scenario analysis rather than included in the base case. The proportion classified as likely non-albicans Candida infection among microscopy-classified acute VVC cases was 13.3% (6/45) in the study cohort. The proportion of colonized women with persistent symptoms warranting selective culture was not directly estimated in this dataset. Because this proportion was not directly estimated in the study dataset, a 10% base-case value was specified as a modeling assumption. A range of 5%–20% was examined in sensitivity analysis. The assumption was informed only broadly by published recurrence estimates (6%–9%) [1] and should not be interpreted as an empirical estimate of persistent symptoms in this cohort.
All women undergoing culture-based diagnosis were assumed to return for a second consultation to receive results, regardless of culture outcome, incurring an additional consultation fee in this pathway only.
The societal-perspective scenario assumed that all women in the culture-based pathway incurred the return-visit transport and time costs. This scenario is not applied to the syndromic management or microscopy-based pathways, which do not require a comparable return visit.
RESULTS
Study Population and Microscopy-based Diagnostic Category Distribution
A total of 447 high vaginal swab records from women evaluated for suspected acute VVC at Sebbi Hospital Laboratory between January 2022 and December 2025 were included in the analysis. The overall median age was 31 years (interquartile range 27–36). The distribution of clinical characteristics and documented symptoms of participants is presented in Table 2.
Table 2.
Symptoms and Clinical Characteristics of participants
| Characteristic | Likely Acute VVC Under the Microscopy-based Classification Framework (N = 45) | Other Microscopy-based Classification (N = 402) |
|---|---|---|
| Age group | ||
| ≤19 | 0 (0.0%) | 12 (3.0%) |
| 20–29 | 18 (40.0%) | 167 (41.5%) |
| 30–39 | 20 (44.4%) | 176 (43.8%) |
| 40–49 | 7 (15.6%) | 37 (9.2%) |
| 50+ | 0 (0.0%) | 10 (2.5%) |
| Pregnancy | ||
| No | 35 (77.8%) | 345 (85.8%) |
| Yes | 10 (22.2%) | 57 (14.2%) |
| Diabetes history | ||
| No | 42 (93.3%) | 382 (95.0%) |
| Yes | 3 (6.7%) | 20 (5.0%) |
| Antibiotic use (last month) | ||
| No | 29 (64.4%) | 323 (80.3%) |
| Yes | 16 (35.6%) | 79 (19.7%) |
| Previous antifungal treatment | ||
| No | 39 (86.7%) | 354 (88.1%) |
| Yes | 6 (13.3%) | 48 (11.9%) |
| Vaginal discharge | ||
| No | 16 (35.6%) | 210 (52.2%) |
| Yes | 29 (64.4%) | 192 (47.8%) |
| Vulvar itching | ||
| No | 23 (51.1%) | 304 (75.6%) |
| Yes | 22 (48.9%) | 98 (24.4%) |
| Vulvar burning | ||
| No | 43 (95.6%) | 398 (99.0%) |
| Yes | 2 (4.4%) | 4 (1.0%) |
| Dyspareunia | ||
| No | 44 (97.8%) | 398 (99.0%) |
| Yes | 1 (2.2%) | 4 (1.0%) |
| Vulvar erythema | ||
| No | 45 (100.0%) | 401 (99.8%) |
| Yes | 0 (0.0%) | 1 (0.2%) |
Data are presented as n (%), where percentages are calculated within Acute VVC and No acute VVC groups (column percentages).
Abbreviations: Acute VVC, combined “likely acute VVC (C. albicans)” and “likely acute VVC (non-albicans Candida species)” microscopy categories; No acute VVC, combined “likely Candida species colonization” and “no laboratory evidence of acute VVC” categories.
Of 447 women evaluated, 39 (8.7%) were classified by the microscopy framework as likely acute VVC with morphology considered consistent with C. albicans, 6 (1.3%) as likely acute VVC with morphology considered compatible with non-albicans Candida infection, 93 (20.8%) as likely Candida species colonization, and 309 (69.1%) as having no laboratory evidence of acute VVC. Forty-five women (10.1%, 95% CI 7.4%–13.2%) were classified as likely acute VVC under the microscopy-based framework, while 402 (89.9%) did not, of whom 309 (76.9%) had no Candida species elements detected and 93 (23.1%) had blastospores without neutrophils and were classified as likely Candida colonization under the microscopy framework.
Clinical Features by Microscopy-based Acute Vulvovaginal Candidiasis Classification
Among clinical features documented at the time of laboratory request, vulvar itching was more frequently documented among women classified as likely acute VVC under the microscopy-based framework, present in 48.9% of women with this classification compared with 24.4% of those without (P < .001). Vaginal discharge and antibiotic use in the preceding month were also more frequently documented among women with acute VVC (64.4% versus 47.8%, P = .034; 35.6% versus 19.7%, P = .013, respectively). Age, pregnancy, diabetes history, previous antifungal treatment, vulvar burning, dyspareunia, and vulvar erythema did not differ significantly between groups (all P > .05). However, given the small number of women classified as likely acute VVC (n = 45), these observations were considered descriptive.
Base Case Cost Consequences
Under syndromic management, all 1000 women were assumed to receive empiric antifungal treatment without laboratory investigation. Total direct cost was USD 10 761 per 1000 women, comprising antifungal treatment costs (USD 1301) and consultation costs (USD 9459), with no laboratory cost incurred (Table 3). In the culture-based diagnosis pathway, 330 women (33.0%) were assumed to have positive culture results and to receive antifungal treatment. Of these, approximately 150 women were estimated to represent overtreatment under the modeled overtreatment fraction of 45.6%, based on the proportion of culture-positive women not classified as likely acute VVC in the secondary analysis. Total direct cost was USD 35 565 per 1000 women, the highest of the 3 pathways, driven primarily by the combined culture service cost for all 1000 women (USD 16 216), antifungal treatment costs (USD 429), and consultation costs (USD 18 919), the latter reflecting two consultations per woman. For the microscopy-based diagnosis, 101 women (10.1%; 95% CI 7.4–%13.2%) classified as likely acute VVC under the microscopy framework were assumed to receive antifungal treatment, while treatment was withheld for the remaining 899 women (89.9%). Total direct cost was USD 14 996 per 1000 women, comprising microscopy costs (USD 5405), antifungal treatment costs (USD 131), and consultation costs (USD 9459).
Table 3.
Cost-Consequence Model Results per 1000 Symptomatic Women-Base Case
| Cost Component | Syndromic Management | Culture-based Diagnosis | Microscopy-based Diagnosis |
|---|---|---|---|
| 1. Laboratory costs | |||
| Women undergoing laboratory testing (n) | 0 | 1000 | 1000 |
| Microscopy unit cost per test | 0 | 0 | 5.41 |
| Culture unit cost per test-USD | 0 | 16.22 | 0 |
| Laboratory cost per woman-USD | 0 | 16.22 | 5.41 |
| Total laboratory cost-USD | 0 | 16 216.22 | 5405.41 |
| 2. Antifungal treatment costs | |||
| Women assigned antifungal treatment (%) | 100% | 33.0% | 10.1% |
| Women assigned antifungal treatment (n) | 1000 | 330 | 101 |
| Model-estimated overtreatment (n) | Not estimated | 150 | Not estimated |
| Women not assigned antifungal treatment (n) | 0 | 670 | 899 |
| Antifungal unit cost per episode-USD | 1.30 | 1.30 | 1.30 |
| Total antifungal treatment cost-USD | 1301.35 | 429.45 | 131.44 |
| 3. Consultation costs | |||
| Weighted mean consultation fee-USD | 9.46 | 9.46 | 9.46 |
| Consultations per woman (n) | 1 | 2 | 1 |
| Total consultation cost-USD | 9459.46 | 18 918.92 | 9459.46 |
| Total direct costs per 1000 women (1 + 2 + 3)-USD | 10 760.81 | 35 564.58 | 14 996.31 |
In a scenario incorporating selective fungal culture within the microscopy-based pathway for women classified as likely non-albicans Candida infection and for women with persistent symptoms following a colonization classification, total direct cost for microscopy-based diagnosis rose to USD 15 552 per 1000 women (range USD 15 383–15 889 across the sensitivity bounds).
Incremental Cost and Break-Even Analysis
At the base case public-sector antifungal prices (USD 1.30 per treatment episode), microscopy-based diagnosis did not generate net direct cost savings when compared with syndromic management. Total direct costs were USD 10 761 under syndromic management and USD 14 996 under microscopy-based diagnosis, an incremental cost of USD 4235 per 1000 women for microscopy-based diagnosis relative to syndromic management. This is because the cost of performing microscopy for all 1000 symptomatic women (USD 5405) exceeded the antifungal treatment savings achieved by withholding treatment from 899 women (USD 1170 saved on antifungals alone).
Culture-based diagnosis exceeded microscopy-based diagnosis by USD 20 568 per 1000 women under the base case and by USD 24 804 per 1000 women compared with syndromic management. This gap widened further under a societal-perspective scenario incorporating patient transport and time costs for the culture pathway's return visit, which raised the cost difference between culture-based and microscopy-based diagnosis to USD 23 464 per 1000 women (range USD 22 834–24 827).
For the base-case microscopy-based pathway, without selective fungal culture, the break-even antifungal unit cost, the price per treated episode at which microscopy-based diagnosis becomes cost-neutral relative to syndromic management, was USD 6.01. It was derived as the microscopy cost per 1000 women (USD 5405) divided by the number of women in whom treatment was avoided (n = 899). This threshold corresponds to UGX 22 237 per episode, approximately 4.6 times the base-case weighted mean antifungal cost of USD 1.30 per episode and 2.4 times the clotrimazole 500 mg pessary dispensing price. When the selective-culture scenario was incorporated, the break-even antifungal unit cost was approximately USD 6.63 per treatment episode under the base-case 10% persistent-culture assumption, with corresponding thresholds of approximately USD 6.59–6.68 across the 5%–20% sensitivity range.
In a scenario analysis incorporating one retreatment course for each woman estimated to represent overtreatment under the culture-based model, total cost of culture-based diagnosis rose from USD 35 565 to USD 35 760 per 1000 women at the base case overtreatment fraction (45.6%), an increase of USD 196. Across the plausible overtreatment range of 20%–70%, this increase varied from USD 86 to USD 301. Across the narrower 95% CI (35.9%–55.7%), it varied from USD 154 to USD 239.
Sensitivity Analysis
The total cost difference between syndromic management and microscopy-based diagnosis was most sensitive to the unit costs of microscopy, followed by the unit cost of antifungal treatment. Variation in the microscopy unit cost produced the largest absolute swing in the cost difference (USD 5405 per 1000 women), followed by antifungal treatment cost (USD 5265). The facility markup applied to antifungal procurement prices produced a smaller change of USD 439 per 1000 women, while the selective-culture scenario changed the cost difference by USD 506 across the 5%–20% range. Variation in the microscopy treatment rate had minimal influence, changing the difference by only USD 75 per 1000 women.
Antifungal treatment cost was the only parameter that reversed the direction of the cost difference (Figure 2A). At USD 0.65 per treatment episode, microscopy-based diagnosis was USD 4820 more costly than syndromic management per 1000 women, whereas at USD 6.51 it was USD 444 less costly. The break-even point occurred at approximately USD 6.01 per treatment episode. In contrast, variation in microscopy unit cost affected only the magnitude of the difference. Reducing the cost to USD 2.70 narrowed the difference to USD 1533 per 1000 women, while increasing it to USD 8.11 widened it to USD 6,938, with syndromic management remaining less costly throughout the range. In the base-case syndromic-versus-microscopy comparison, the culture positivity rate, routine fungal culture unit cost, consultation fee, and overtreatment fraction did not affect the cost difference because they did not enter either pathway's base-case cost structure.
Figure 2.
One-way sensitivity analysis of direct cost differences between diagnostic pathways per 1000 symptomatic women. Each bar shows the range of the cost difference resulting from varying one parameter across its specified plausible range while holding all other parameters at their base-case values. The white circle indicates the base-case value, the red dashed vertical line indicates the base-case cost difference, and the black vertical line at zero indicates equal costs between the two pathways. A, Compares syndromic management with the base-case microscopy-based diagnosis pathway. The base-case cost difference was USD −4235, calculated as syndromic management cost minus microscopy-based diagnosis cost. The negative value therefore indicates that microscopy-based diagnosis was more costly under base-case assumptions. Only variation in the antifungal unit cost crossed the zero-cost threshold. Selective fungal culture was examined separately as a scenario analysis and was not included in the base-case cost difference or the USD 6.01 break-even threshold. B, Compares culture-based diagnosis with microscopy-based diagnosis. The base-case cost difference was USD 20 568. The culture-based pathway includes a second consultation for all women to receive culture results and a treatment decision. All parameter ranges remained above zero, indicating that culture-based diagnosis was more costly than microscopy-based diagnosis under every parameter combination examined. The unit cost of fungal culture was the dominant parameter, followed by the consultation fee, reflecting the two consultations assumed for the culture-based pathway compared with one for microscopy-based diagnosis.
For the comparison between culture-based and microscopy-based diagnosis, the fungal culture unit cost was the dominant parameter, producing a swing of USD 16 216 per 1000 women (Figure 2B). The consultation fee was the second most influential parameter, with a swing of USD 8108, reflecting the 2 consultations assumed for the culture-based pathway compared with 1 for microscopy-based diagnosis. Variation in microscopy unit cost produced a swing of USD 5405, while antifungal treatment cost produced a smaller swing of USD 1341. The selective-culture scenario changed the cost difference by USD 506, whereas culture positivity rate, antifungal facility markup, and microscopy treatment rate produced smaller changes of USD 130, USD 112, and USD 75, respectively. Culture-based diagnosis remained more costly than microscopy-based diagnosis across all parameter combinations, with the cost difference ranging from approximately USD 12 460 to USD 28 676 per 1000 women.
Overtreatment was examined separately because, when retreatment costs were excluded, it did not affect the base-case cost difference. If each woman classified as overtreated under culture-based diagnosis received one additional antifungal course, the additional cost was USD 196 per 1000 women at the base-case overtreatment fraction of 45.6%. Across the 95% CI of 35.9%–55.7%, the additional cost ranged from USD 154 to USD 239 per 1000 women. Across the wider plausibility range of 20%–70%, it ranged from USD 86 to USD 301 per 1000 women.
DISCUSSION
This study estimated and compared the direct medical costs of three diagnostic pathways for acute VVC in a routine laboratory setting in Uganda, using empirical unit costs and a 10.1% model-assigned treatment proportion derived from 447 consecutive laboratory assessments. At base-case public-sector antifungal prices, culture-based diagnosis was the most costly diagnostic pathway for acute VVC, exceeding syndromic management and microscopy-based diagnosis by USD 24 804 and USD 20 568 per 1000 women, respectively, and remained more costly across all parameter ranges and scenarios examined. At base case scenario, microscopy-based diagnosis did not generate net direct cost savings over syndromic management, as the cost of performing microscopy for all symptomatic women exceeded the savings from avoided antifungal treatment at current public-sector prices. However, the absence of direct medical cost savings with microscopy-based diagnosis when compared with syndromic management should not be interpreted as evidence against its clinical use. Rather, the findings indicate that the economic value of microscopy depends on the relationship between diagnostic costs and the cost of antifungal treatment avoided, as well as on the consequences of more selective treatment.
Only the unit price of antifungal treatment reversed the direction of the cost difference between microscopy-based diagnosis and syndromic management. For the base-case microscopy pathway, without selective fungal culture, the break-even treatment price was USD 6.01 per episode, above which microscopy became less costly than empiric treatment. The selective-culture scenario increased the total cost of the microscopy-based pathway and was therefore analyzed separately.
Nearly 90% of women in this routine laboratory dataset did not meet the microscopy criteria for acute VVC requiring antifungal treatment. This finding is important when interpreting the economic implications of syndromic management. Although withholding treatment from these women did not offset the cost of universal microscopy at current public-sector antifungal prices, the microscopy-based framework substantially changed the proportion of symptomatic women assigned antifungal treatment under the model. The potential value of this more selective treatment approach therefore extends beyond immediate drug expenditure. Assignment of antifungal treatment to a smaller proportion of women may have implications for antimicrobial stewardship and for reducing repeated or inappropriate treatment, although these downstream consequences were not directly measured in this study. The previously reported economic burden of recurrent VVC in Uganda [2] further highlights the importance of distinguishing active infection from other causes of vulvovaginal symptoms, although the present study cannot establish that the diagnostic pathway itself accounts for that broader burden.
The cost disadvantage of culture-based diagnosis was robust across all sensitivity ranges and was driven primarily by the cost of fungal culture. The consultation fee was a secondary contributor because the model assumed two consultations for the culture pathway: an initial consultation for specimen collection and a return consultation for results and treatment where indicated. In contrast, syndromic management and microscopy-based diagnosis were modeled with a single consultation because they provide a treatment decision without a separate result insurance visit. These suggest that the economic disadvantage of routine culture-based diagnosis arises not only from the laboratory test itself, but also from the additional healthcare encounter required under the model assumptions.
The selective use of culture within the microscopy-based diagnosis produced only a modest increase in total cost, with the base-case microscopy cost increasing from USD 14 996 to USD 15 552 per 1000 women. This scenario was designed to reflect selective investigation of likely non-albicans VVC cases and women with persistent symptoms after initial assessment rather than routine culture of all symptomatic women. Within the modeled scenarios, selective culture was associated with a smaller additional cost than routine culture of all symptomatic women and may therefore be a more economically plausible strategy when additional species-level or microbiological information is expected to alter management. This is consistent with the role assigned to culture in clinical guidances [5, 11].
The findings also illustrate the importance of distinguishing the cost of a diagnostic pathway from its broader clinical value. Syndromic management was least costly under the base-case public-sector prices, but it assumed empiric treatment of all symptomatic women. Microscopy required an additional laboratory expenditure but identified a much smaller group for whom antifungal treatment was indicated. Culture provides additional microbiological information but at substantially greater laboratory and consultation costs. Consequently, the choice among these pathways should not be based on direct treatment expenditure alone, particularly where the objectives include obtaining additional diagnostic information and limiting empiric antifungal exposure.
Limitations
Several limitations warrant acknowledgement. The microscopy-based classification framework was not independently validated against fungal culture, molecular testing, or another reference standard in the 447-record dataset. The distinction between likely acute VVC and Candida colonization was based on prespecified interpretations of observed Candida morphology and inflammatory cells and therefore may be subject to misclassification. If it occurs, misclassification could alter the estimated proportion assigned antifungal treatment and, consequently, the modeled treatment costs and incremental cost differences between diagnostic pathways. The findings presented should therefore be interpreted as the economic consequences of the specified model-based classification framework and its assumptions, rather than as evidence of a microscopy-definitively-established infection or colonization status.
The base-case analysis adopted a healthcare-system perspective. Patient transport and productivity costs were incorporated only in a separate societal-perspective scenario for the culture pathway's return visit and were not modeled across all 3 pathways. The societal analysis therefore represents a scenario rather than a complete societal-perspective cost analysis. The antifungal unit cost was derived from public-sector procurement prices with a 60% facility-level markup [15]. Because this parameter was the most influential determinant of the syndromic-versus-microscopy comparison, uncertainty in the actual price paid by patients or facilities could materially alter the economic conclusion. The break-even threshold should therefore be interpreted as a decision threshold rather than as evidence that microscopy is universally cost saving.
The culture positivity rate was derived from a published sub-Saharan African meta-analysis [8] rather than a local empirical estimate, although the parameter was varied across its reported 95% CI. Additionally, the overtreatment fraction was derived from a validation study in pregnant women and may not be directly generalizable to the broader population represented by the present laboratory dataset. This parameter was therefore examined separately through a retreatment scenario, including both its reported 95% CI and a wider 20%–70% plausibility range. Even under these assumptions, the additional retreatment cost remained relatively small compared with the overall cost difference between culture-based and microscopy-based diagnosis.
The proportion of colonized women with persistent symptoms who would warrant selective fungal culture was not directly estimated from the dataset. The scenario therefore used a 10% base-case assumption with a 5%–20% sensitivity range. Direct empirical estimates of this parameter in Ugandan clinical populations would improve future modeling. Finally, retreatment following diagnostic failure was excluded from the base case and examined only as a separate scenario. One additional treatment course for each woman classified as overtreated under culture-based diagnosis increased pathway costs by USD 196 per 1000 women at the base-case overtreatment fraction, and by USD 86–301 across the wider plausibility range. These analyses indicate that retreatment does not materially change the principal cost ranking of the 3 pathways under the assumptions used.
Despite these limitations, this study provides locally grounded evidence on the economic consequences of alternative diagnostic approaches to acute VVC in Uganda. The results indicate that microscopy-based diagnosis is substantially less costly than routine culture-based diagnosis and provides a mechanism for more selective antifungal treatment than syndromic management. However, at current public-sector antifungal prices, microscopy does not produce a net direct medical cost saving because the cost of universal microscopy exceeds the immediate drug savings. Its potential value therefore lies in the provision of additional microscopy-based diagnostic information, more selective assignment of antifungal treatment under the model, and the potential implications for antifungal stewardship. These benefits should be considered alongside the direct costs when evaluating the role of microscopy in routine VVC care.
CONCLUSIONS
At current public-sector antifungal prices in Uganda, the microscopy-based pathway did not generate net direct medical cost savings compared with empiric syndromic management because the cost of performing microscopy for all symptomatic women exceeded the immediate savings from reduced model-assigned antifungal treatment. For the base-case microscopy pathway, without selective fungal culture, the break-even antifungal price was USD 6.01 per treatment episode, indicating that the economic comparison could change where treatment is obtained at higher prices.
Routine culture-based diagnosis was consistently the most costly of the three modeled pathways, exceeding microscopy-based diagnosis by USD 20 568 per 1000 women under base-case assumptions and remaining more costly across the parameter ranges examined. Its higher cost was driven principally by the fungal culture unit cost and the additional consultation required for the return visit.
Under the specified microscopy-based classification framework, 89.9% of women in the routine laboratory dataset were not assigned antifungal treatment. This indicates that microscopy-based classification may provide a more selective basis for treatment decisions and may have implications for antifungal stewardship, although diagnostic accuracy, clinical outcomes, recurrence, and antifungal resistance were not measured. The results support further evaluation of microscopy-based strategies, with fungal culture used selectively when additional microbiological information is clinically required, rather than demonstrating that microscopy is clinically superior or should universally serve as the primary diagnostic approach.
Supplementary Material
Notes
Acknowledgments. We thank Sebbi Hospital Management, Sebbi Hospital Laboratory Department Staff, Margret Nyilabakunzi, Flavia Dhikhosoka, Charity Mutesi, Julius Okwi, Kabiito Duncan Matovu, Moses Oketch, Shakira Babirye, and Joy Happy Banonya.
Author Contributions. J. B.: Conceptualization, methodology, software, formal analysis, data curation, visualization, writing—original draft, project administration. N. P. P.: Methodology, data curation. G. M.: Conceptualization, methodology, formal analysis, writing—review and editing, supervision. A. K.: Methodology, software, A. W. W.: Methodology, software, formal analysis, writing—review and editing. F. O.: Methodology, software, supervision. A. R. A.: Methodology, software, funding, writing—review and editing, supervision. B. K. N.: Conceptualization, methodology, formal analysis, writing—review and editing, supervision. All authors approved the final manuscript.
Data availability. All data generated or analyzed during this study are included in this published article and its supplementary data.
Financial support. This study was supported by the Bill and Melinda Gates Foundation, through Uganda National Institute of Public Health (UNIPH), grant number INV-059606. The content of the article is the sole responsibility of the authors and does not necessarily reflect the official views of the funding source.
Contributor Information
Jonani Bwambale, Uganda National Institute of Public Health (UNIPH), Ministry of Health, Kampala, Uganda; Clinical Laboratory, Sebbi Hospital, Nansana Municipality, Wakiso District, Uganda; Karis Medical Clinic, Kampala, Uganda.
Nakitende Patricia Peace, Clinical Laboratory, Sebbi Hospital, Nansana Municipality, Wakiso District, Uganda.
Ssebunya Enock, Karis Medical Clinic, Kampala, Uganda.
Gerald Mboowa, The African Centre of Excellence in Bioinformatics and Data-Intensive Sciences, Infectious Diseases Institute, Makerere University, Kampala, Uganda.
Anita Kisakye, Uganda National Institute of Public Health (UNIPH), Ministry of Health, Kampala, Uganda.
Abel W Walekhwa, IDEMU Mathematical Modelling Unit, Department of Epidemiology and Biostatistics, Makerere University School of Public Health, Kampala, Uganda.
Felix Ocom, Uganda National Institute of Public Health (UNIPH), Ministry of Health, Kampala, Uganda.
Alex R Ario, Uganda National Institute of Public Health (UNIPH), Ministry of Health, Kampala, Uganda.
Betty Kivumbi Nannyonga, Department of Mathematics, School of Physical Sciences, College of Natural Sciences, Makerere University, Kampala, Uganda.
Supplementary Data
Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
References
- 1. Rautemaa-Richardson R, Sobel JD, Stone N, et al. State-of-the-art review: managing vulvovaginal candidiasis. Clin Infect Dis 2026; 82:371–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Jonani B, Bongomin F. Economic burden of recurrent vulvovaginal candidiasis in Uganda: a cost-of-illness analysis. IJID Reg 2025; 16:100680. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Omosa-Manyonyi GS, de Kam M, Tostmann A, et al. Evaluation and optimization of the syndromic management of female genital tract infections in Nairobi, Kenya. BMC Infect Dis 2023; 23:547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Mendonça A, Santos H, Franco-Duarte R, Sampaio P. Fungal infections diagnosis – past, present and future. Res Microbiol 2022; 173:103915. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. CDC . Testing and diagnosis for candidiasis. 2024. Available at: https://www.cdc.gov/candidiasis/testing/index.html (Accessed 2 May 2026).
- 6. Cornely OA, Sprute R, Bassetti M, et al. Global guideline for the diagnosis and management of candidiasis: an initiative of the ECMM in cooperation with ISHAM and ASM. Lancet Infect Dis 2025; 25:e280–93. [DOI] [PubMed] [Google Scholar]
- 7. Halliday C, Alguacil-Cuéllar L, Chen SC-A, Alastruey-Izquierdo A. Optimizing antifungal therapies for Candida infections: evidence, resistance, and emerging approaches. Clin Microbiol Infect 2026; 32:365–73. [DOI] [PubMed] [Google Scholar]
- 8. Mushi MF, Olum R, Bongomin F. Prevalence, antifungal susceptibility and etiology of vulvovaginal candidiasis in sub–Saharan Africa: a systematic review with meta-analysis and meta-regression. Med Mycol 2022; 60:myac037. [DOI] [PubMed] [Google Scholar]
- 9. Willems HME, Ahmed SS, Liu J, Xu Z, Peters BM. Vulvovaginal candidiasis: a current understanding and burning questions. J Fungi (Basel) 2020; 6:27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Vieira-Baptista P, Grincevičienė Š, Oliveira C, Fonseca-Moutinho J, Cherey F, Stockdale CK. The international society for the study of vulvovaginal disease vaginal wet mount microscopy guidelines: how to perform, applications, and interpretation. J Low Genit Tract Dis 2021; 25:172–80. [DOI] [PubMed] [Google Scholar]
- 11. Saxon Lead Author Guideline Development Group Cara; Edwards A, Rautemaa-Richardson R, Owen C, et al. British Association for Sexual Health and HIV national guideline for the management of vulvovaginal candidiasis (2019). Int J STD AIDS 2020; 31:1124–44. [DOI] [PubMed] [Google Scholar]
- 12. Timmermans B, De Las Peñas A, Castaño I, Van Dijck P. Adhesins in Candida glabrata. J Fungi (Basel) 2018; 4:60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Shiri T, Birungi J, Garrib AV, et al. Patient and health provider costs of integrated HIV, diabetes and hypertension ambulatory health services in low-income settings—an empirical socio-economic cohort study in Tanzania and Uganda. BMC Med 2021; 19:230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Uganda Bureau of Statistics . National labour force survey 2021 main report. Kampala, Uganda: Uganda Bureau of Statistics, 2022. [Google Scholar]
- 15. Medicine Transparency Alliance . Medicine price components in Uganda. Kampala, Uganda: Dumaic Global Health, 2015. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.