The first aim of the study was to develop a new measure of slum-like living conditions based on the Uganda Demographic and Health Surveys (UDHS) data. In line with the definition of slums provided by UN-Habitat [15], we identified five key domains that capture slum-like conditions, which can be derived from national household surveys: substandard housing conditions, limited water access, overcrowding, unclean cooking fuel, and limited toilet access. In terms of substandard housing conditions and hygiene, indicators such as the absence of improved toilets, roofs, walls, and floors were considered. Limited access to clean water was characterized by longer travel times to water sources and the absence of water within the premises. Overcrowding was measured by households occupying single rooms and having at least five individuals per sleeping room. Furthermore, the use of unclean fuel for cooking and the sharing of a single toilet by at least four households were identified as key indicators of unhealthy cooking materials and limited toilet access, respectively. These conditions have been consistently associated with adverse child health outcomes, the occurrence and recurrence of infections, and accidents [8, 28]. For instance, houses constructed with poor-quality materials such as substandard wood, thatch, and cardboard pose a higher risk of fire accidents [29]. Overcrowding and limited access to water are associated with the transmission of infectious diseases [8, 28].

Based on the UN-Habitat definition of slum conditions, a significant proportion of the population in GKMA, approximately 97%, resided in households characterized by at least one slum-like condition between 2006 and 2016. This definition, while comprehensive, has limitations as it categorizes most low- and lower-middle-income countries as slum areas, making it challenging to capture modest improvements in living conditions. There is a need for more flexible measures that can effectively monitor progress in these countries. It is important to recognize that living standards vary greatly both within and between countries [22], rendering a globally standardized measure less suitable for accurately estimating local situations. Additionally, the concept of deprivation implies a standard of living below that of the majority in a specific context [22], which contradicts the UN-Habitat approach. To address these measurement challenges, we developed the slum-like severity score based on standardized scores, identifying those living in severely deprived conditions as individuals whose standardized scores fell below the average standard of living conditions.

Based on our localized measure, we discovered that as of 2016, households in GKMA had an average slum-like score of 30%. This indicates that the majority of households in this region experience at least 3 out of 10 measures of slum-like living conditions. Moreover, there have been no discernible improvements in these conditions over the years.

Throughout the years, sharing a toilet with at least five households, living in a one-room household, and lacking water within the premises were the primary contributors to slum-like deprivation. At the same time, there has been an increase in the number of households without a separate place or room for cooking. These findings emphasize the prevalent issues related to poor water and sanitation conditions, highlighting the urgent need for interventions that improve water and sanitation infrastructure within urban communities to alleviate slum-like conditions.

Our estimates of the population living in slum-like conditions are in close agreement with the 2014 estimates provided by UN-Habitat [30] and slightly higher than the figures reported in national reports [17]. As of 2016, around 67% of households in GKMA were characterized by at least 30% of slum-like conditions, and there has been no significant change in this percentage over the span of a decade (66% in 2006 and 70% in 2011). Furthermore, approximately 31% of households were classified as having severely slum-like conditions. The likelihood of living in a slum-like household was higher in the neighboring city outskirts within the overall GKMA than within the city of Kampala itself. This trend can be attributed to urban expansion and inadequate urban planning, particularly in these neighboring districts where construction activities are taking place without sufficient public infrastructure and adherence to building regulations [18].

The second objective of our study was to examine the characteristics of caregivers of children living in slum-like conditions. Consistent with previous research [31, 32], we found that lower levels of education, being young in age, being employed in unprofessional jobs, and being married were associated with a higher risk of residing in households with greater slum-like conditions. These findings shed light on the socio-economic factors that influence health outcomes in urban areas. In cities, young individuals often migrate to slum areas where they may experience early pregnancies or marriages [33]. Moreover, while urban settings offer employment opportunities, those living in poverty often face limited skills and end up in low-paying jobs. Consequently, they tend to reside in slum-like households as that is what they can afford, even though these conditions are typically characterized by substandard living conditions.

Lastly, we aimed to investigate the association between slum-like living conditions and child nutrition, common illnesses, and healthcare-seeking behavior in GKMA. We identified disparities in the prevalence of common illnesses and malnutrition among children living in different levels of slum-like conditions. The likelihood of experiencing fever, anemia, diarrhea, and stunting significantly increased with the severity of slum-like conditions. These findings align with existing research that links these conditions to nutritional deficiencies, infections, and poor environmental circumstances [28, 29], which were identified as key indicators of slum-like living conditions in our study. Furthermore, previous studies focusing on intra-urban health disparities have shown that the likelihood of stunting and anemia is higher among children living in poorer urban clusters compared to those in non-poor urban clusters [34,35,36]. These results are not surprising, as more affluent families generally have greater access to resources that support better childcare practices, thus protecting their children from infections and related health issues.

The residents of slum-like households in GKMA exhibited low treatment and care-seeking behavior when it came to fever among under-5 children. Surprisingly, however, there were no differences in accessing treatment and care for suspected diarrhea based on slum-like status. Various factors, such as maternal education and employment status, have been identified as influential in mothers’ healthcare-seeking behavior. However, the impact of employment status on healthcare-seeking can be complex. In an urban context, women engaged in full-time employment face the challenge of balancing family and work responsibilities, which limits their time to seek healthcare services. This is especially true for those in the informal sector who may have no time or leave days to attend to their sick children. Additionally, caregivers or mothers often seek treatment based on the symptoms of the illness. Our findings indicate the need for child health interventions that focus on preventive and protective measures for all urban residents, which could help to reduce the need to seek out and obtain treatment. However, further studies are required to understand the barriers that prevent access to appropriate healthcare services.

The study has a major limitation in that it only used household-based items to construct the Slum-like Severity Index, overlooking other community context indicators such as population density, uncollected garbage, contaminated water, open sewers, and poor drainage, which are clear indicators of slum areas. However, the items included in the analysis align with the specific criteria outlined by UN-Habitat in its definition of slum or slum-like living conditions. Additionally, the DHS dataset is not specifically designed to provide precise estimates for slum and non-slum dwellers. Nevertheless, by leveraging the DHS GIS data and national administrative GIS boundary data, we were able to reconstruct the DHS cluster/district names, which improved the sample size sufficient to measure slum-like housing conditions. Another limitation is that the reporting of fever and diarrhea in the UDHS relies on self-reporting, which involves recalling past events and self-definition of the severity of conditions. This may introduce some reliability issues in the estimates. However, the use of a reference period of collecting data on events that occurred in the last two weeks prior to the survey helps to minimize recall biases.

In conclusion, our analysis of national household surveys has provided estimates of households experiencing slum-like conditions in the Greater Kampala Metropolitan Area. The study results enhance our understanding of the precarious living conditions in urban areas and offer insights into the characteristics of individuals living in such conditions. The results indicate that a significant proportion of households in the region continue to face poor household living conditions, with little change observed over a decade. The identified indicators linked to slum-like conditions include the sharing of toilets among at least five households, inadequate water access within premises, and residing in single rooms. These conditions have notable implications for childhood health issues, nutrition, and the utilization of healthcare services for children. They also highlight the presence of health inequalities within urban areas. Therefore, efforts should be focused on improving housing quality, ensuring access to clean water and sanitation facilities, and enhancing healthcare services to alleviate the health challenges faced by populations living in slum-like conditions. Furthermore, various individual factors such as age, employment type, education level, and marital status have emerged as key determinants of living in slum-like households. Recognizing these factors is crucial for addressing the social determinants of health in urban settings and tailoring appropriate healthcare services.

Additionally, it is important to conduct larger survey samples to gain a deeper understanding of the living conditions experienced by disadvantaged households in urban areas and their impact on child health and nutrition. By doing so, we can further enhance our knowledge and inform targeted interventions to improve the well-being of these vulnerable populations.