Microbiological Non-Compliance with Escherichia coli and Salmonella spp. in Prepared Foods in Relation to the Hygienic and Sanitary Conditions of Restaurants in Independencia, Lima, Peru

1. Introduction

Food safety remains a major global public health challenge. The World Health Organization (WHO) estimated that more than 866 million cases of foodborne illness and approximately 1.52 million deaths occurred worldwide in 2021, according to estimates published in 2026 based on global foodborne disease burden data for 2000–2021 [1]. Contaminated food can transmit hazards associated with more than 200 diseases and generate healthcare costs, economic losses, and adverse effects on trade and tourism [1]. In this context, the WHO Global Strategy for Food Safety 2022–2030 emphasizes strengthening surveillance systems, verifying regulatory compliance, improving laboratory capacity, and promoting risk-based prevention [2].

Prepared foods served in restaurants are of particular importance because they may pass through multiple stages at which contamination can occur, including raw material reception, storage, washing, cutting, thawing, heat treatment, post-cooking handling, environmental exposure, and serving. The Codex Alimentarius General Principles of Food Hygiene establish that good hygiene practices should operate in conjunction with preventive systems designed to control hazards, particularly when foods are intended for consumption without further treatment capable of reducing the microbial load [3].

The microbiological quality of prepared foods is therefore the outcome of multiple processes occurring both within and outside the establishment. Da Cunha [4] notes that food safety practices in food service operations depend on the interaction of knowledge, behavior, organizational factors, and environmental conditions. Similarly, Rifat et al. [5] and Negassa et al. [6] indicate that safe food-handling practices are associated not only with training but also with infrastructure, the availability of supplies, and working conditions.

Among the microorganisms used to assess food safety, Escherichia coli requires particular interpretive consideration. The detection of generic E. coli or concentrations exceeding microbiological criteria may serve as an indicator of inadequate hygienic practices and potential fecal contamination; however, its identification does not automatically demonstrate the presence of a pathogenic strain. Establishing whether an isolate belongs to STEC, ETEC, EPEC, or another pathotype requires specific characterization of virulence factors. This distinction is essential to avoid overstating the clinical significance of a microbiological finding that is used primarily as an indicator of sanitary quality.

The relevance of E. coli in the food chain has been extensively documented. Babines-Orozco et al. [7], in a review focusing on Mexico and Latin America, reported that food-associated strains may exhibit diverse antimicrobial resistance profiles and highlighted the importance of strengthening their surveillance. Foxcroft et al. [8], in their analysis of microbiological trends in ready-to-eat foods from restaurants, cafés, and catering establishments, identified non-STEC E. coli as a continuing relevant indicator of contamination in these settings.

Salmonella spp., by contrast, has a different significance. It is a recognized causative agent of salmonellosis and foodborne outbreaks. Its presence in prepared foods intended for consumption constitutes a finding of public health significance, even when the proportion of positive samples is low. The European One Health zoonoses report has consistently identified salmonellosis among the most frequently reported zoonoses and highlighted the persistence of foodborne outbreaks in settings with well-developed health surveillance systems [9].

In Latin America, Guerrero et al. [10] conducted a systematic review of nontyphoidal Salmonella in foods from 13 countries and found substantial variability in prevalence, serotypes, and antimicrobial resistance profiles. The considerable heterogeneity observed indicates that microbiological risk depends on the food matrix, production practices, and the specific sanitary conditions of each context. The authors also emphasized that Salmonella control requires an integrated approach across the food chain.

This perspective is particularly important because contamination detected in restaurants does not necessarily originate exclusively within the establishment. Thomas et al. [11] described several pathways through which Salmonella and pathogenic E. coli can be introduced into fruits and vegetables during pre- and postharvest stages, including through soil, irrigation water, animals, storage, preparation, and handling. Bhatia et al. [12] reached a similar conclusion in their review of microbiological risks associated with ready-to-eat fresh produce, highlighting the contributions of water, transportation, storage, processing, and food handlers to final product contamination.

Nevertheless, restaurant operations can either amplify or mitigate risks introduced through raw materials. Bukhari et al. [13], who analyzed 294 samples from restaurants in Mecca, detected microbiological contamination on surfaces and utensils used during food preparation. Similarly, Tareq et al. [14] identified fecal contamination by E. coli on food handlers’ hands and food-contact surfaces in restaurants, highlighting the potential for cross-contamination among the food preparation environment, food handlers, and food products.

Ready-to-eat foods are a particularly relevant food matrix because they do not always undergo further heat treatment before consumption. Through a systematic review and meta-analysis of studies conducted in developing countries, Mengistu et al. [15] estimated pooled prevalences of 23.8% for E. coli and 17.4% for Salmonella in ready-to-eat foods. However, the authors observed considerable heterogeneity across the included studies, reinforcing the need to interpret these findings in light of the characteristics of the food and the establishment.

Studies examining specific food matrices further confirm this heterogeneity. Łepecka et al. [16], in their assessment of ready-to-eat salads, detected E. coli in a high proportion of samples and also identified Salmonella spp. In contrast, McLauchlin et al. [17], who analyzed 604 salad products obtained from retail outlets and catering establishments in England, detected no Salmonella, although some samples exhibited unsatisfactory levels of E. coli.

Castrica et al. [18], after analyzing thousands of ready-to-eat food samples from retail outlets and canteens in northern Italy, reported an overall Salmonella prevalence of only 0.07%, although the microorganism was detected more frequently in preparations containing raw ingredients. This finding demonstrates that low prevalence rates are possible in settings with more robust control measures, while also highlighting the continued significance of occasional detection of the microorganism.

In the context of prepared foods served in Latin American restaurants, Guadalupe-Moyano et al. [19] analyzed rice, fish, juices, and salads served in restaurants in San Pablo, Ecuador. The authors detected neither E. coli nor Salmonella spp.; however, they identified non-compliance with coliform criteria and incomplete adherence to good practices, demonstrating that the absence of these microorganisms during a sampling event does not necessarily indicate the absence of other hygiene deficiencies.

Marquis et al. [20] evaluated ceviche, poke, and sushi dishes sold at retail establishments in California. Among the 105 samples analyzed, generic E. coli was detected in a small proportion of samples, whereas Salmonella was not detected. The relevance of this study to Latin American contexts lies particularly in ceviche, a product that undergoes extensive handling and is frequently consumed without conventional heat treatment.

In other contexts, the proportions may be considerably higher. Tonjo et al. [21] in their study of raw meat intended for ready-to-eat consumption from hotels and restaurants in Ethiopia, identified bacterial contamination and antimicrobial resistance profiles indicative of substantial food safety deficiencies. Differences across studies show that the type of preparation influences microbiological risk and that the term “prepared foods” encompasses food matrices with markedly different microbiological characteristics.

Barreira et al. [22], in their study of ready-to-eat street foods in Lisbon, likewise demonstrated that microbiological quality can vary even within the same urban setting. This evidence highlights the value of jointly assessing specific microorganisms and food preparation conditions, rather than inferring food safety solely from the type of establishment.

Even formal food safety assurance systems do not guarantee the complete absence of contamination. Subramaniam et al. [23] identified microbiological non-compliance in ready-to-eat foods and on surfaces in HACCP-certified establishments in Malaysia. Lai et al. [24] documented microbiological hygiene issues in certified restaurants and hotel kitchens in eastern China, while Osaili et al. [25] found that restaurants accounted for the largest number of microbiologically unsatisfactory surfaces among the 308 establishments assessed in Dubai.

This evidence supports the need to complement sanitary inspections with microbiological results. Barnes et al. [26] highlight that inspections constitute an internationally recognized tool for public health protection, although their methodologies and effectiveness require continuous evaluation. Kosola et al. [27] observed that certain non-compliances identified during inspections could precede foodborne outbreaks, whereas Saarimaa et al. [28] demonstrated that targeted control interventions can influence subsequent compliance outcomes.

Kim et al. [29] also showed that certain characteristics of inspection programs are associated with the prevention of restaurant-associated foodborne outbreaks. However, Levy et al. [30], in a systematic review and meta-analysis, cautioned that although food safety policies are associated with reductions in microbial contamination, the available evidence is characterized by substantial heterogeneity and important methodological limitations. Therefore, a decrease observed between two observational visits cannot automatically be interpreted as a causal effect of the inspection.

In Latin America, Nascimento and Lima [31] described the persistence of hygienic and sanitary deficiencies in Brazilian fast-food establishments, while Ribeiro et al. [32] identified areas for improvement in the infrastructure, facilities, equipment, and utensils of popular restaurants. This evidence shows that the existence of sanitary regulations does not automatically ensure consistent compliance across all components of an establishment.

From a methodological perspective, Bulochova et al. [33] found that much of the research on food safety in food service operations relies on questionnaires, observations, or microbiological analyses conducted separately and recommended triangulating different types of measurements. Similarly, Mota et al. [34] noted that combining environmental and product monitoring provides a better understanding of potential sources and pathways of contamination.

In Peru, Ministerial Resolution No. 591-2008/MINSA establishes the microbiological criteria applicable to food and beverages for human consumption, while Technical Health Standard No. 142-MINSA/2018/DIGESA regulates the sanitary conditions of restaurants and related services [35,36].

Although the Peruvian regulatory framework establishes microbiological criteria for food and sanitary requirements for restaurants, an empirical gap remains regarding the extent to which hygienic and sanitary classifications assigned during inspections can identify establishments where microbiological non-compliance is concentrated. Integrating the sanitary assessment of establishments with microbiological analysis of prepared foods is relevant for determining the extent to which observable hygiene conditions correspond to the microbiological quality of products intended for consumers.

In this context, the present study provides descriptive evidence from two assessments conducted in the same 170 restaurants in Independencia, Lima, by examining the distribution of microbiological non-compliance involving Escherichia coli and Salmonella spp. according to the hygienic and sanitary classification of the establishments. The findings may help inform risk-based microbiological surveillance strategies, without assuming that sanitary classification can replace food testing or that the observed associations establish causality.

Therefore, this study aimed to analyze microbiological non-compliance involving Escherichia coli and Salmonella spp. in prepared foods according to the hygienic and sanitary conditions of restaurants in the district of Independencia, Lima, Peru, across two assessments conducted between 2022 and 2024.

2. Materials and Methods

2.1. Study Design and Setting

A quantitative, observational, non-experimental study was conducted, involving two microbiological and sanitary assessments of the establishments. The study was carried out in restaurants in the district of Independencia, Lima Province, Lima Department, Peru, during the period 2022–2024.

The restaurant was the unit of analysis, and one prepared-food sample was collected from each establishment during each visit.

2.2. Population and Sample

The reference population consisted of 304 registered restaurants in the district of Independencia, Lima, Peru. A total of 170 establishments located across different geographical areas and restaurant types were assessed. However, the available documentation did not allow the establishment selection procedure to be reliably reconstructed or the use of probability or non-probability sampling to be established. Therefore, no sampling method that could not be substantiated by the original records was retrospectively assigned.

Each establishment was assessed on two occasions, and one prepared-food sample was collected during each visit. A total of 170 samples were obtained per assessment, yielding 340 sample records across both rounds. Because the selection procedure could not be verified, the results are interpreted in relation to the establishments assessed, and their statistical representativeness of the 304 restaurants registered in the district is not assumed.

2.3. Assessment of Hygienic and Sanitary Conditions

Hygienic and sanitary conditions (HSC) were assessed using an inspection checklist developed based on Technical Health Standard No. 142-MINSA/2018/DIGESA, Sanitary Standard for Restaurants and Related Services.

The assessment considered conditions related to kitchen facilities, storage, sanitary facilities, food handling, cleaning and disinfection, cross-contamination prevention, waste management, temperature control, and other sanitary aspects of the establishment.

Based on the overall compliance percentage, the establishments were classified as unhealthy (0–50%), in progress (51–74%), or healthy (75–100%).

2.4. Sampling of Prepared Foods

At each visit, one prepared-food sample was collected from each restaurant. Each sample was individually labeled before transport to the laboratory.

The samples were transported in insulated containers with cooling gel packs, with efforts made to maintain a temperature below 10 °C. Microbiological processing was performed within 12 hours of sample collection.

2.5. Microbiological Analyses

The quantification of Escherichia coli was expressed as the most probable number per gram (MPN/g). Results were classified as compliant or non-compliant by comparison with the microbiological criteria established in Ministerial Resolution No. 591-2008/MINSA, taking into account the category applicable to the food analyzed.

For the detection of Salmonella spp., qualitative presence-or-absence results were recorded and interpreted according to the microbiological criterion applicable to the food category assessed. The final description of the analytical procedures should specify the method used for each microorganism, including the corresponding technical reference, in accordance with the laboratory’s original records.

Molecular typing of E. coli was not performed, and virulence genes were not analyzed. Therefore, the results were interpreted as indicators of microbiological non-compliance rather than as confirmation of the presence of specific enteropathogenic pathotypes.

2.6. Statistical Analysis

Absolute frequencies and percentages were calculated for samples that failed to meet the microbiological criteria in each of the two assessments. Non-compliance was also described according to the hygienic and sanitary classification of the restaurant.

The association between the hygienic and sanitary classification and non-compliance with the microbiological criteria for E. coli was assessed separately for each evaluation using Spearman’s rank correlation coefficient. For Salmonella spp., given the small number of positive results, the findings were described according to the hygienic and sanitary classification, without conducting statistical inferences regarding the association between the variables.

Statistical significance was set at p < 0.05. Differences in the observed frequencies between the two assessments were examined descriptively. No statistical tests for paired data were applied to formally assess changes between assessments in the same study units.

2.7. Ethical Considerations

The information collected during the hygienic and sanitary assessments and microbiological analyses was treated confidentially, and the results are presented in aggregate form, without individually identifying the establishments assessed. The documentation available for preparing the manuscript did not allow verification of whether approval had been granted by an ethics committee or a formal exemption had been issued. Therefore, no institutional authorization is reported unless its existence can be substantiated through documentary evidence. The need for ethical review and, where applicable, the relevant approval or exemption must be verified with the responsible institution before publication of the study.

3. Results

3.1. Microbiological Non-Compliance in Prepared Foods

During the first visit, 23 of the 170 prepared-food samples exceeded the microbiological criterion established for E. coli, corresponding to 13.53%. At the second assessment, only one non-compliant sample was recorded (0.59%). This represented a descriptive reduction of 22 samples, equivalent to a 95.65% decrease relative to the initial frequency.

For Salmonella spp., five samples tested positive during the first assessment (2.94%). No positive samples were detected during the second visit (0/170). This finding applies exclusively to the samples analyzed during that assessment and does not demonstrate the permanent absence of the microorganism.

Table 1. Microbiological Non-Compliance Involving E. coli and Salmonella spp. in Prepared Food.

Microorganism	First visit, n (%)	Second visit, n (%)	Absolute change
E. coli	23 (13.53)	1 (0.59)	−22
Salmonella spp.	5 (2.94)	0 (0.00)	−5

Note. n = 170 samples per visit. The values represent samples that exceeded the corresponding microbiological criteria.

Table 1. Microbiological Non-Compliance Involving E. coli and Salmonella spp. in Prepared Food.

Microorganism	First visit, n (%)	Second visit, n (%)	Absolute change
E. coli	23 (13.53)	1 (0.59)	−22
Salmonella spp.	5 (2.94)	0 (0.00)	−5

Note. n = 170 samples per visit. The values represent samples that exceeded the corresponding microbiological criteria.

3.2. Microbiological Non-Compliance According to Hygienic and Sanitary Conditions

The distribution of cases showed a clear concentration in establishments with unfavorable sanitary classifications. During the first assessment, of the 23 samples that failed to meet the microbiological criterion for E. coli, 22 came from restaurants classified as unhealthy, one from a restaurant classified as in progress, and none from restaurants classified as healthy. During the second assessment, the only non-compliant result was recorded in a restaurant classified as unhealthy. For Salmonella spp., all five positive samples from the first assessment came from restaurants classified as unhealthy; no cases were recorded in restaurants classified as in progress or healthy. During the second assessment, no positive results were detected in any category.

Table 2. Distribution of Non-Compliant Samples According to Hygienic and Sanitary Classification.

Microorganism	Assessment	Unhealthy	In Progress	Healthy	Total
E. coli	First	22	1	0	23
E. coli	Second	1	0	0	1
Salmonella spp.	First	5	0	0	5
Salmonella spp.	Second	0	0	0	0

Table 2. Distribution of Non-Compliant Samples According to Hygienic and Sanitary Classification.

Microorganism	Assessment	Unhealthy	In Progress	Healthy	Total
E. coli	First	22	1	0	23
E. coli	Second	1	0	0	1
Salmonella spp.	First	5	0	0	5
Salmonella spp.	Second	0	0	0	0

Of the 28 microorganism-specific non-compliance records in the first assessment, 27 corresponded to restaurants classified as unhealthy. This count represents microbiological determinations and not necessarily distinct samples or establishments, as a single sample may have been analyzed for both microorganisms.

3.3. Association Between HSC and Escherichia coli

During the first assessment, a weak but statistically significant inverse association was observed between the hygienic and sanitary classification and non-compliance with the microbiological criterion for E. coli:

ρ = −0.296; p < 0.001.

This indicates that more favorable sanitary classifications tended to be associated with fewer samples failing to meet the microbiological criterion.

During the second assessment, the following result was obtained:

ρ = −0.104; p = 0.177.

The association remained negative but was weak and did not reach statistical significance. The presence of only one non-compliant sample during the second assessment limited the variability available for this analysis.

Table 3. Spearman’s Correlation Between Hygienic and Sanitary Classification and E. coli Non-Compliance.

Microorganism	First Assessment		Second Assessment
	ρ	p	ρ	p
E. coli	−0.296	<0.001	−0.104	0.177

Note. n = 170 restaurants per assessment. The coefficients represent the association between the hygienic and sanitary classification and E. coli non-compliance.

Table 3. Spearman’s Correlation Between Hygienic and Sanitary Classification and E. coli Non-Compliance.

Microorganism	First Assessment		Second Assessment
	ρ	p	ρ	p
E. coli	−0.296	<0.001	−0.104	0.177

Note. n = 170 restaurants per assessment. The coefficients represent the association between the hygienic and sanitary classification and E. coli non-compliance.

4. Discussion

The main finding of this study was that microbiological non-compliance involving E. coli and Salmonella spp. in prepared foods was concentrated predominantly in restaurants classified as unhealthy. During the first assessment, 22 of the 23 cases of non-compliance involving E. coli occurred in this category, while all five samples positive for Salmonella spp. also originated from unhealthy establishments. During the second assessment, only one sample failed to meet the microbiological criterion for E. coli, and no Salmonella spp. was detected. These findings indicate a correspondence between unfavorable sanitary conditions and a higher frequency of adverse microbiological results.

The initial proportion of non-compliance for E. coli was 13.53%, lower than the 23.8% estimated by Mengistu et al. [15] in their meta-analysis of ready-to-eat foods in developing countries. However, this comparison should be interpreted with caution because the review included heterogeneous food types and methodologies, with substantial variability across studies. This heterogeneity indicates that microbiological contamination depends on the type of food preparation, the origin of ingredients, handling practices, storage, sanitary conditions, and the laboratory criteria applied.

Foxcroft et al. [8] also showed that non-STEC E. coli remains a relevant microorganism in the monitoring of ready-to-eat foods from restaurants, cafés, and catering establishments. The primary significance of generic E. coli lies in its role as an indicator of inadequate hygienic practices, rather than in automatically assuming that all isolates possess pathogenic potential.

This distinction is particularly important in the present study. The microbiological analysis did not include molecular typing or screening for virulence genes. Consequently, the 23 initial cases of non-compliance should not be described as “23 foods contaminated with pathogenic E. coli,” but rather as 23 samples that exceeded the established microbiological criterion for E. coli. The review by Babines-Orozco et al. [7] shows that food-associated E. coli strains in Latin America exhibit considerable epidemiological diversity and variation in antimicrobial resistance profiles, supporting the need for future molecular analyses.

The concentration of 22 of the 23 initial non-compliant results in restaurants classified as unhealthy provides an additional finding. Although the correlation coefficient indicated a weak association (ρ = −0.296), the descriptive distribution suggests that microbiological non-compliance was not randomly distributed across sanitary categories. Restaurants with more favorable hygienic and sanitary conditions had fewer adverse microbiological events.

This pattern is biologically plausible. Hygienic and sanitary conditions encompass several components, including cleaning, disinfection, separation of raw and cooked foods, temperature control, food handling, and environmental sanitation. Any of these factors may influence the likelihood that microorganisms initially present in raw ingredients survive, multiply, or are transferred to prepared foods. The Codex Alimentarius principles specifically emphasize the importance of integrated control of the stages at which microbiological hazards may be introduced or persist.

Thomas et al. [11] highlight that enteric microorganism can reach food during stages preceding its arrival at the restaurant, including through soil, irrigation water, animals, and storage. Bhatia et al. [12] further note that transportation, handling, and food handlers subsequently influence the overall risk. Therefore, the association observed in Lima does not imply that all contamination originated within the restaurant; hygienic and sanitary conditions may influence both the introduction of new microorganisms and the establishment’s ability to control risks already present in raw materials.

Studies of food-contact surfaces support this potential transmission pathway. Bukhari et al. [13] found microbiological contamination on various utensils and surfaces in restaurants, whereas Tareq et al. [14] detected fecal contamination by E. coli on hands, tables, plates, and utensils. These findings demonstrate that cross-contamination can link different components of the food preparation environment to the finished food product.

Regarding Salmonella spp., the initial proportion of 2.94% was considerably lower than the pooled estimates reported in some meta-analyses of studies conducted in developing countries. Mengistu et al. [15] reported a pooled prevalence of 17.4%, whereas Guerrero et al. [10] documented an extremely wide range of prevalence estimates for nontyphoidal Salmonella in Latin American foods. These differences do not diminish the significance of the findings from Lima, since even a low frequency of Salmonella in foods intended for consumption has public health implications.

Ma et al. [37], in a meta-analysis of fresh fruits and vegetables sold at retail, estimated a pooled Salmonella prevalence of approximately 3%, a value similar to the 2.94% initially observed in restaurants in Independencia. This comparison is useful only as an epidemiological reference because the food matrices evaluated were different.

The distribution of Salmonella spp. also showed a descriptive concentration of positive results in establishments with unfavorable sanitary conditions: all five positive samples from the first assessment originated from restaurants classified as unhealthy. However, given the small number of events, this pattern should be interpreted descriptively and does not, by itself, support a robust inferential association.

During the second assessment, Salmonella spp. was not detected in any of the 170 samples. The absence of detection during this round represents a favorable finding for the samples analyzed; however, it does not demonstrate that the microorganism has been permanently eliminated from the restaurants.

The literature clearly illustrates this distinction. Castrica et al. [18], in more than 6,000 determinations of Salmonella in ready-to-eat foods, found an overall prevalence of only 0.07%. Even in systems with very low levels of contamination, positive findings continued to be recorded, particularly in foods containing raw ingredients.

McLauchlin et al. [17], in their analysis of 604 salad products collected in England, did not detect Salmonella and reported only a small proportion of unsatisfactory results for E. coli. This study demonstrates that the absence of Salmonella detection during a sampling round is possible, but it should not be interpreted as evidence of the microorganism’s permanent absence.

Similar findings were reported by Guadalupe-Moyano et al. [19] in restaurants in Ecuador. The authors detected neither E. coli nor Salmonella spp. in the preparations analyzed, although they identified non-compliance with other microbiological indicators and partial adherence to good practices. This evidence is particularly relevant because it demonstrates that the absence of two specific microorganisms does not necessarily indicate a fully hygienic food service system.

Marquis et al. [20] provide another point of comparison. In ceviche, poke, and sushi sold at retail establishments in the United States, they detected generic E. coli in a small proportion of samples and found no Salmonella. The authors emphasized the importance of microbiological surveillance of raw ready-to-eat foods because these products lack a final heat-treatment step capable of eliminating contamination.

In contrast, Łepecka et al. [16] reported a less favorable scenario in ready-to-eat salads, with a substantial presence of E. coli and the detection of Salmonella. These findings demonstrate that the microbiological quality of prepared foods depends substantially on their composition, processing, storage, and handling.

Tonjo et al. [21] also found high levels of contamination in raw meat intended for ready-to-eat consumption served in hotels and restaurants. Although this food matrix has a microbiological profile that differs from that of the range of preparations examined in Lima, these findings show that microbiological risk can increase considerably when food does not undergo an adequate inactivation step.

Barreira et al. [22], in their analysis of ready-to-eat street foods, found variability in microbiological quality, reinforcing the need to assess the foods served to consumers rather than focusing exclusively on the facilities.

The findings of the present study should also be considered alongside evidence from formally regulated establishments. Subramaniam et al. [23] documented microbiological contamination in HACCP-certified facilities, whereas Lai et al. [24] identified microbiological non-compliance in restaurants and hotel kitchens, although they did not detect E. coli in the cooked foods examined. This confirms that formalized procedures may reduce risks but do not replace objective verification.

Osaili et al. [25] reached a similar conclusion after analyzing 947 samples from 308 establishments in Dubai: 87% met microbiological quality standards, but restaurants accounted for the largest number of microbiologically unsatisfactory surfaces. Therefore, sanitary controls should simultaneously consider the final food product and the environmental components that may contribute to its contamination.

In the restaurants in Independencia, the decrease in E. coli non-compliance from 23 samples to one, together with the subsequent absence of detected Salmonella, coincided temporally with a second sanitary assessment. However, this temporal sequence does not allow the conclusion that the inspection was the sole cause of the decrease.

Levy et al. [30] found that food safety interventions in catering establishments were associated with an overall reduction of approximately 28.6% in microbial counts, but highlighted the methodological heterogeneity of the available evidence. Barnes et al. [26] also noted that inspections are widely used as a protective measure, although their effectiveness should be evaluated systematically.

Kosola et al. [27] demonstrated that certain non-compliances identified during inspections may precede foodborne outbreaks, whereas Saarimaa et al. [28] found that targeted interventions can influence subsequent compliance outcomes. Similarly, Kim et al. [29] identified characteristics of inspection programs associated with the prevention of restaurant-associated foodborne outbreaks.

These findings suggest that sanitary surveillance may contribute to risk control, but they do not justify interpreting the decrease observed in Lima as a “proven effect” of the visits. There was no control group of restaurants that had not undergone inspection, no random assignment, and no standardized intervention. Furthermore, factors such as changes in suppliers, the types of food served, seasonality, staff turnover, demand, or voluntary changes implemented by the establishments may have contributed to the observed outcome.

Another relevant aspect is the loss of statistical significance in the association between hygienic and sanitary conditions (HSC) and E. coli during the second assessment. The coefficient changed from −0.296, with p < 0.001, to −0.104, with p = 0.177. This does not necessarily mean that sanitary conditions ceased to be associated with food safety; rather, only one microbiological event was observed among the 170 observations, sharply reducing the variability available for testing the association.

Bulochova et al. [33] highlight the need to use complementary measurements and methods capable of adequately characterizing risk in food service operations. In the present study, Spearman’s rank correlation coefficient was used to describe the monotonic relationship between sanitary classification and non-compliance involving E. coli; however, this coefficient does not specifically account for the binary nature of the outcome or the dependence between the two assessments of each restaurant. Future studies should apply methods designed for binary outcomes and repeated measurements.

Monitoring the final product also complements environmental assessment. Mota et al. [34] emphasize that microbiological surveillance programs are more informative when they integrate results from food and the surrounding environment. This approach makes it possible to determine whether observable hygiene conditions correspond to actual microbiological findings.

From a Latin American perspective, Nascimento and Lima [31] documented persistent hygienic and sanitary deficiencies in fast-food establishments in Brazil, while Ribeiro et al. [32] identified areas for improvement related to facilities, equipment, and utensils in popular restaurants. The evidence obtained in Lima complements these studies by linking the sanitary conditions of establishments with two specific microbiological outcomes in prepared foods.

From an operational perspective, establishments classified as unhealthy should be prioritized for microbiological food sampling. This classification can guide risk-based surveillance but should not replace microbiological testing of the product.

The scientific contribution of this study lies in providing evidence from two assessments of the same restaurants regarding the correspondence between hygienic and sanitary classification and the microbiological results of prepared foods. In the first assessment, 22 of the 23 samples that were non-compliant for E. coli and all five samples positive for Salmonella spp. came from restaurants classified as unhealthy. This convergence supports the potential use of sanitary classification to prioritize sampling; however, it does not validate its predictive capacity or, by itself, demonstrate a causal relationship. Its discriminatory performance should be evaluated in future studies using specific analyses.

Therefore, microbiological analysis of prepared foods provides information that an exclusively visual or documentary inspection cannot provide. The sanitary inspection checklist helps identify potential risk conditions, whereas laboratory analysis determines whether the product meets the applicable microbiological criteria. Integrating these two sources of information provides a more comprehensive basis for guiding sanitary surveillance.

5. Conclusions

Microbiological non-compliance involving Escherichia coli and Salmonella spp. in prepared foods was concentrated primarily in restaurants with unfavorable hygienic and sanitary conditions. During the first assessment, 22 of the 23 non-compliant results for E. coli corresponded to establishments classified as unhealthy, and one corresponded to the in-progress category, whereas all five positive results for Salmonella spp. originated exclusively from restaurants classified as unhealthy. None of these initial non-compliance findings were recorded in establishments classified as healthy.

The frequency of non-compliance for E. coli decreased descriptively from 23 samples (13.53%) to one sample (0.59%) between the two assessments. For Salmonella spp., the number of positive samples decreased from five (2.94%) to no detections during the second visit. These results indicate a marked reduction in microbiological findings, although they do not support the conclusion that the microorganisms were permanently eliminated from the establishments.

Hygienic and sanitary conditions showed an inverse and statistically significant association with E. coli non-compliance during the first assessment, whereas the association did not reach statistical significance during the second assessment, when only one non-compliant sample was recorded. For Salmonella spp., the five positive results from the first assessment and the absence of detection during the second assessment were interpreted descriptively because of the small number of events.

Therefore, the findings support an integrated assessment of hygienic and sanitary conditions and the microbiological quality of prepared foods. This approach may help identify establishments that should be prioritized for surveillance, objectively verify product compliance, and guide corrective measures toward restaurants where non-compliance is concentrated. Differences observed between visits should be interpreted as descriptive changes rather than as demonstrated causal effects of the inspections.

5.1. Study Limitations

The study was observational and did not include a control group; therefore, the decrease in microbiological findings cannot be causally attributed to the sanitary inspections. In addition, the selection procedure for the 170 restaurants was not sufficiently documented to confirm a probability sampling design, limiting the generalizability of the findings to the 304 establishments registered in the district.

Furthermore, prepared foods constituted a heterogeneous category, and the available aggregated data do not allow robust comparisons by specific dish type, main ingredient, or heat treatment.

The identification of E. coli did not include pathotype characterization or the analysis of virulence genes; therefore, the results should be interpreted as microbiological non-compliance involving E. coli and not as confirmation of enteropathogenic strains.

Molecular characterization, serotyping, and antimicrobial resistance analysis of Salmonella spp. were also not performed.

The low frequency of Salmonella and the occurrence of only one E. coli non-compliant result during the second assessment substantially limited the statistical power of the inferential analyses.

Spearman’s rank correlation does not specifically account for the binary nature of microbiological non-compliance or the dependence between the two measurements obtained from each restaurant. Furthermore, no paired statistical test was applied to formally assess changes between visits; therefore, the temporal comparison is descriptive, and the inferential results should be interpreted with caution.

5.2. Implications and Future Research Directions

Sanitary surveillance could use the hygienic and sanitary classification as a criterion for prioritizing microbiological food sampling, particularly in restaurants classified as unhealthy.

Future studies should individually document the type of food, raw and cooked ingredients, cooking temperature, storage conditions, time elapsed between preparation and serving, source of raw materials, and handling practices after heat treatment.

Molecular characterization of the isolates is also recommended, including the identification of E. coli pathotypes, Salmonella serotypes, virulence genes, and antimicrobial resistance profiles.

Using the original individual-level dataset, the association should be reanalyzed using models for binary outcomes and repeated measures, such as mixed-effects logistic regression or generalized estimating equations. When events are extremely rare or complete separation occurs, penalized or exact methods should be considered.