Background

Tuberculosis (TB) remains one of the leading infectious causes of death globally, with an estimated 10.6 million cases and 1.3 million deaths in 2022, second only to COVID-19 [1]. Transmission is highest when TB is unrecognized and untreated, as there is a sharp drop in infectiousness after only a few days of effective treatment [2], highlighting the importance of prompt diagnosis and treatment initiation. Diagnostic delays lead to more severe disease, higher infectiousness [3], increased development of drug-resistance [4], poor treatment outcomes [5, 6], and higher transmission [7, 8].

Delays in the diagnosis and treatment initiation have been reported worldwide. A systematic review of 58 studies from both high- and low/middle-income countries (LMICs) reported median diagnostic delays ranging from 60 to 90 days, with the highest delays (> 120 days) seen in countries with high TB prevalence, owing to factors related to both the individuals seeking healthcare and the healthcare systems [9]. A review of exclusively LMICs found median delays from 30 to up to 367 days between symptom onset and TB diagnosis [10].

Peru is a middle-income country with the highest estimated TB incidence in Latin America (151 cases per 100,000 population) and one of the highest burdens of multidrug-resistant TB globally [11]. In 2015, a median diagnostic delay of 8 weeks was reported by a study in the capital Lima, driven by multiple factors, including poor awareness of the disease, the cost of care, and lack of social support [12]. Though such barriers to seeking care for TB have been well described [9, 10, 12, 13], less is known about the obstacles faced by individuals once connected to the healthcare system. A comprehensive understanding of the factors that influence the time to TB diagnosis is crucial to designing effective strategies to reduce its morbidity and mortality. With this study, we aimed to measure the TB diagnostic delay in a low-resource district of Peru and examine the experiences people with TB had navigating the healthcare system, with the goal of identifying areas for potential intervention.

Methods

Study design and setting

We conducted a sequential explanatory mixed methods study to understand barriers to prompt TB diagnosis in Lima, Peru and identify opportunities for improvement. Our study took place in Carabayllo district of Lima, which has a population of over 400,000, predominantly of middle and low income [14]. Public sector health facilities operated by the Ministry of Health include health posts (basic primary care centers without X-ray services), health centers (larger primary care centers, some of which have laboratories and X-ray services), and regional referral hospitals [15]. The private sector offers clinics, laboratories, and private hospitals. Diagnosis is based primarily on sputum smear microscopy, chest radiography, and clinical evaluation. Diagnostics and treatment are free in the public sector. Some private sector health facilities provide TB diagnostic services, but none provide treatment. In the public health system, passive TB screening is done for individuals presenting with respiratory symptoms, with active screening reserved for those with a known exposure to a person with TB. An active TB screening program including mobile units that offer portable X-rays and sputum testing with GeneXpert MTB/RIF has been implemented in Carabayllo District since 2019 by the non-profit non-governmental organization Socios En Salud, in collaboration with the Ministry of Health [16]. People who are diagnosed with TB in these units are referred to the public health facilities for further evaluation and treatment.

Quantitative data collection and analysis

We enrolled a convenience sample of 100 adults (≥ 18 years old) who started TB treatment during November 2020-April 2021 and October-February 2022 at the 12 Ministry of Health primary care facilities in Carabayllo (enrollment was paused in between because of the COVID-19 pandemic). We purposefully recruited participants to ensure distribution across the different health facilities and a mix of pulmonary and extrapulmonary TB diagnoses. Study staff would regularly check clinic registers and recruit people who had recently initiated treatment, aiming to complete the interview within one month of treatment initiation to improve recall. We administered structured surveys asking about dates of symptom onset and each encounter with the healthcare system until the diagnosis that enabled the person to get treatment. We calculated total diagnostic delay (time from symptom onset to diagnosis), delay before contact with the health system (time from symptom onset until the first visit to a health facility or mobile TB screening unit; pharmacies were not included as they cannot diagnose TB) and delay after contact with the health system (time from the first visit to a health facility until diagnosis). We describe these values in weeks since recalled dates were sometimes not accurate to a specific day, but participants could generally recall the week of the month.

Qualitative data collection and analysis

We categorized delays both before and after first contact with the health system as < 4 weeks and ≥ 4 weeks. We recruited interview participants purposefully to encompass a range of short and long delays before and after contact with the health system, with representation of all possible combinations of these delays and of age groups and sexes. One-time one-on-one interviews lasting approximately 1 h were conducted in Spanish in the health center or the participant’s home during April–May 2021 and November 2021-January 2022. The study team member who administered the survey recruited interview participants from among those who had agreed to be recruited, and the same study team member performed the interview. All study team members conducting interviews (DA, EA, HC, SF, GM, JR, IT) were female Peruvian nurse technicians who had no relationship with the interviewee prior to the study and who were trained in interviewing. The interview guide asked the participant to recount their experience obtaining a diagnosis from the time of symptom onset and to identify factors that helped or hindered their ability to be diagnosed. Audio recordings of interviews were transcribed by a local Peruvian transcriber and checked by the interviewer for fidelity; transcripts were not shared with participants. We originally planned to enroll 30 participants for in-depth interviews, but thematic saturation was reached after 26 interviews. Twelve individuals declined interview participation upon recruitment, and none withdrew during the interview.

CG reviewed audio recordings and coded written transcripts using the qualitative data management software Dedoose (version 9.0.17). Emerging themes were identified through open coding using an inductive approach. Concepts related to diagnostic barriers and facilitators were identified from the participants’ narratives of their own experience and their responses to interviewer prompts. CG and CMY discussed and resolved cases of code drift and redundancy. Identified concepts were compiled into a codebook and organized into main thematic categories.

Integration

We used a joint display [17] to map themes from the qualitative interviews onto the participant pathway described by the quantitative analysis, identifying factors that affected delays before and after the first contact with the health system.

Results

Quantifying delays in diagnosis

The characteristics of individuals who participated in surveys and interviews is shown in Table 1. The median total diagnostic delay was 9 weeks (interquartile range [IQR] 4–22 weeks). The median delay before contact with the health system was 4 weeks (IQR 0–9 weeks) and the median delay after contact with the health system was 3 weeks (IQR 0–9 weeks). Over a third (37%) of participants had their first healthcare encounter after more than 4 weeks of symptoms, and 35% received the diagnosis of TB more than 4 weeks after their first encounter. Overall, 29% of participants first sought care at a Ministry of Health primary care facility, 9% at a Ministry of Health hospital, 42% in the private sector, and the remaining 20% in other locations such as a mobile screening unit or government health facility not operated by the Ministry of Health. Participants made a median of 3 (IQR 2–4) health system visits to obtain a diagnosis, and 51% reported having to miss time from work or arrange childcare to attend at least one visit. At the first visit, 60% of individuals reported that they received an X-ray or sputum test, and 14% reported being diagnosed with TB.

Table 1 Characteristics of individuals with tuberculosis recruited to participate in surveys (N = 100) and in-depth interviews (N = 26)
Full size table

Barriers and facilitators to tuberculosis diagnosis in Lima, Peru

A total of 73 emerging concepts related to barriers and facilitators to TB diagnosis were identified and organized into 12 main thematic categories using a framework of healthcare seeker-related and healthcare system-related factors. The joint display in Fig. 1 illustrates how the qualitative themes map to the quantitative results. Examples of barriers were more prominent than facilitators in participants’ narratives. Only one of the main themes (social and family support, 1.F.) consistently referred to a facilitator, while some other themes emerged as contrasting examples of barriers and facilitators (e.g. awareness of TB as a facilitator and lack of awareness as a barrier). Therefore, in the qualitative results below we briefly summarize the facilitators to draw attention to them, as most of the elaboration focuses on barriers.

Fig. 1
figure 1

Factors contributing to delays in the diagnosis of TB. Abbreviations: IQR, interquartile range. TB, tuberculosis

Full size image
  1. 1.

    Healthcare seeker factors

Six main themes related to patient-centered factors impacting the time to TB diagnosis emerged: knowledge and awareness of TB, misconceptions about the disease, attitudes and practices around health and healthcare, disease presentation, economic factors, and social and family support (Fig. 1). Facilitators in this domain included being acquainted with someone who had previously had TB, having supportive friends or relatives, and having access to free services.

  1. A

    Knowledge and awareness of TB

Some participants expressed having never heard of the disease or having had little to no knowledge of how it manifests. Other patients were acquainted with a relative or friend who had had TB, and in one case the patient participant had previously had the disease himself. They were therefore able to identify their symptoms more promptly and seek medical care. Participants attributed their symptoms to other causes, including upper respiratory infections such as COVID-19, as well as other medical conditions such as asthma, diabetes, or depression.

“I started to have a cough. A simple cough, I thought it was a cold. My throat would close up. I was speaking like I had a closed throat, nothing else. Because I didn’t have a fever, I didn’t have any other symptoms. (…) Yes, I was losing weight, but I thought it was because of diabetes.”

Female participant with pulmonary TB (Interview 21)

  1. B

    Misconceptions about the disease

The notion that TB is associated with a lack of self-care was common among participants. Poor nutrition, lack of sleep and emotional stress were common theories for why one might develop the disease. Due to such notions, some participants believed they were not susceptible to infection. This perceived lack of susceptibility was identified as an important reason why men and younger people may have poor engagement with the health system, leading to delayed diagnosis.

“Well, this is a disease that attacks the lungs especially. If you don’t treat it in time, it can get more complicated in the lungs… you can improve your lifestyle, because in this regard, my life was in disarray. I wasn’t eating at the right times, perhaps. For example, I was playing sports, but I wasn’t eating as I should.”

Male participant with extrapulmonary TB (Interview 10).

Some individuals, despite knowing of the disease, had misconceptions about its presentation. Some believed cough to be a necessary symptom of TB, while one participant did not suspect TB because she lacked hemoptysis. People who were ultimately diagnosed with extrapulmonary TB had believed that the disease could only affect the respiratory system.

  1. C

    Attitudes and practices around health and healthcare

People living with TB often decided to postpone seeking care because they lacked the time, minimized their symptoms, distrusted the medical system, or were afraid of the potential diagnosis. Some expressed they had feared being judged or excluded by others. Indeed, shame and fear of stigmatization was a factor often cited by participants as a reason why someone might not seek medical attention.

“I was fearful because I know relatives who have gone to the hospital and were not treated well. On top of that, in the news, you see that people are not treated well, and they die. You see many things and it gives you fear.”

Female participant with pulmonary TB (Interview 15)

Self-medication was a common practice when participants presumed their illness be a common cold or the flu, and sometimes at the advice of friends and family. When seeking care for the first time, some participants initially went to pharmacies, where pharmacy technicians gave them empiric treatment for alternative diagnoses without a recommendation for further specialized medical evaluation.

  1. D

    Disease presentation

Participants reported having had symptoms not typically associated with TB, which led to a delayed recognition of the disease. Participants who were ultimately diagnosed with extrapulmonary TB presented with atypical symptoms such as abdominal pain, back pain, headaches and, in one case seizures. Even among participants with pulmonary TB, atypical symptoms such as sore throat drove them and their providers to consider upper respiratory infections as the most likely cause of their symptoms. In some cases, the individual’s symptoms waxed and waned or improved with treatments for alternative diagnoses, which deterred them from seeking further care.

  1. E

    Economic factors

The high costs of medical visits, diagnostic tests and transportation posed a significant strain on participants and their families. Even if they were able to afford it themselves, participants identified this as an important reason why others might not be able to receive adequate medical attention. Indeed, one participant who lacked health insurance was not able to afford diagnostic testing and had her care interrupted as a result. In such cases, having access to free services such as free X-rays helped expedite care.

“What happens is that there are also people who don’t have the means to have an X-ray. I was lucky that they did my X-ray for free, because precisely that day [the mobile screening unit] was at the door of the health post (…) And it was free, and that expedited the nurse being able to help me with the treatment. I didn’t have to wait anymore, I didn’t have to spend, because I was already economically devastated.”

Male participant with extrapulmonary TB (Interview 11)

  1. F

    Social and family support

Friends and relatives were often first to recognize the severity of the participant’s symptoms and to encourage them to seek care. Some participants only sought help when their symptoms became severe enough that alarmed friends or relatives brought them to a health facility. In several cases, the participant’s family was instrumental in providing transportation to medical appointments, helping to cover the costs of services, and advocating for their health.

“My parents help me, and I worked, even in the pandemic, I also started working. I saved what I earned. And with that money I went to get care. (…) It saved me that my dad supported me with what I didn’t have.”

Female participant with pulmonary TB (Interview 8)

  1. 2.

    Healthcare system factors

Themes related to the healthcare system that impacted the time to TB diagnosis were organized into six main categories: multiple healthcare encounters, provider interpersonal skills, provider specialty, management decisions, unavailability of services, and waiting times (Fig. 1). Facilitators in this domain included being evaluated by a pulmonologist and having a formal referral document when seeing multiple providers.

  1. A

    Multiple health encounters

Almost all participants described having multiple encounters with the healthcare system prior to receiving the diagnosis of TB, making this the dominant and overarching theme of the narratives. Although all participants were recruited from public sector facilities (where they received free TB treatment), they described using both public and private sectors, sometimes alternating between the two. In some cases, multiple encounters were caused by providers initially favoring non-TB diagnoses. When participants’ symptoms did not improve after being given treatment for an non-TB condition, they sought care again from a different provider. However, multiple encounters were also necessary once a TB evaluation was started, as participants had to schedule subsequent visits to obtain diagnostic tests or receive test results. Challenges included the unavailability of diagnostic tests, long wait times, general practitioners deferring to a specialist, and the fact that TB treatment can only be prescribed in the public sector. Given that X-rays are not consistently available across public health facilities, some participants went to private clinics to obtain them, but clinicians in private clinics who suspected TB referred participants back to public health centers for final diagnosis and treatment.

During the multiple encounters required for the TB diagnostic process, participants were made responsible for their own care in multiple ways. They were tasked with bringing sputum samples for testing, bringing their test results to providers, and in one case an individual who underwent a lymph node biopsy was asked to bring the sample to the pathology laboratory himself. Participants who were directed to a different provider or health facility were often asked to arrange the transfer process themselves, particularly if moving from the private to the public sector.

“He told me, ‘Well now we’re going to refer you to a TB program, but we don’t do it - the [private] clinic doesn’t do it. So you have to go to the health post where you live and ask about the TB program that the government provides, and they will see to your treatment. So from here there is nothing more for us to do. Bye, champ.’”

Male participant with extrapulmonary TB (Interview 16)

Referral documentation was a facilitator to prompt diagnosis. When being referred to another facility for medical evaluation or diagnostic testing, having a prescription or referral document facilitated and expedited the process. For a particular individual living with human deficiency virus (HIV), a referral specifying his condition allowed them to be prioritized in line for a health center.

  1. B

    Provider interpersonal skills

Some participants complained about the attitudes and behavior of individual health professionals, with many stating that the provider was impolite, refused to see them, or did not dedicate enough time to listening to them, ultimately deciding to seek care elsewhere.

“I had to leave the [public] health post, because the way I was treated by the doctor was not right. She even yelled at me. She yelled at me badly and I had to leave. So I sort of lost hope in being able to return to the health post, and I went to a [private] polyclinic.”

Male participant with extrapulmonary TB (Interview 11)

  1. C

    Provider specialty

For some participants, the point at which they received a diagnosis was when they were evaluated by a pulmonologist, often in the private sector. Participants felt that the ability to see a specialist was what expedited their diagnosis and considered the insufficient number of pulmonologists in the public health centers to be a barrier to diagnosis.

“I went to a [private] polyclinic to see pulmonology. They saw the X-ray that I had gotten first, and when he saw my X-ray he asked me to get a tomography. I got the tomography and the same day the doctor also saw it and detected that it was TB.”

Female participant with pulmonary TB (Interview 18)

  1. D

    Management decisions

Providers made decisions that led to delays in the participants’ diagnoses, whether by pursuing other diagnoses, failing to pursue TB diagnostic tests, or failing to follow up. Participants reported providers pursuing other diagnoses because they failed to recognize the symptoms of TB or attributed the symptoms to alternative conditions. In some cases, participants might have had more than one concomitant condition. Some participants perceived their provider as lacking knowledge about the disease. Some described general practitioners who felt uncomfortable pursuing routine TB diagnostics and who advised the participant to seek care from a pulmonologist elsewhere, delaying the diagnostic work-up and ultimately the diagnosis.

“[The doctor] told me ‘you don’t have the necessary symptoms to have the screening test.’ But I did have the necessary symptoms to have other types of tests for TB – fever, pain in my lymph nodes. So instead of helping me, giving me any advice, she sent me directly home to receive a treatment that wasn’t even appropriate, because she had not told me what it was.”

Male participant with extrapulmonary TB (Interview 16)

  1. E

    Inaccessibility of services

Services were not always available. Participants reported seeking care but finding that the health center was closed, the healthcare professional was absent, or diagnostic tests such as X-rays were lacking. The hours of operation of health centers also posed a challenge – as most operate during normal business hours, participants reported conflicts with their work schedule. This was identified as an important factor why people with TB might delay seeking care. Women felt this was especially the case for men, who tend to have longer work hours. When asked about other potential barriers to care, participants mentioned the location of healthcare centers as being hard to reach for populations living in remote and less privileged areas, with some suggesting that services should be brought to these communities.

“(…) in our health post there is no pulmonologist. There should be someone who can take X-rays and take bloodwork also, because in [the health post] there is no place to get bloodwork. Or to get an X-ray - there isn’t one.”

Female participant with pulmonary TB (Interview 4).

  1. F

    Waiting times

Participants recounted waiting in line to see a provider or undergo a test. For one individual, the long wait times deterred them from seeking care at a public hospital altogether. Some participants waited for several days or weeks to obtain their test results, especially those who had biopsies done and were ultimately diagnosed with extrapulmonary TB. In cases where tests were readily available, including sputum microscopy that was quickly read as positive, individuals were promptly directed towards next steps.

“I decided to go to a private doctor because I was working and I could not miss an entire day of work. The private doctor gave me an evening appointment and saw me. (…) You know that in the [public] hospital it is like you said - you have to wait in line, you have to get an appointment, they do not see you then, and it is a bit more difficult than with a private doctor.”

Female participant with pulmonary TB (Interview 21)

Discussion

We observed delays in TB diagnosis resulting from factors related to both people with TB and the healthcare system. Our finding of a median delay of nine weeks from symptom onset to diagnosis is consistent with delays reported in other countries [9, 10], as well as a previous study in Peru [12]. Delays in going to a health facility were caused by the inconvenience of public health facility hours for people who work, self-medication, and lack of perceived susceptibility to TB, often due to lack of knowledge about the disease. Delays after individuals entered the health system were caused by the necessity for multiple healthcare encounters to reach a diagnosis. Factors leading to multiple visits included the inaccessibility of diagnostic services, long wait times, providers’ failure to consider TB or lack of confidence in diagnosing it, the absence of a standardized referral process, and the lack of integration between the public and private sectors. Together, our findings suggest that better strategies for community education, accessibility of health resources, provider training, and integration between the private and public sectors should be prioritized.

As described in other studies, though participants had heard of TB, there was poor awareness and frequent misconceptions about its mode of transmission and presentation [9, 10, 12]. Many participants did not consider that they could fall ill with the disease and often did not seek care until the symptoms became persistent or severe. A previous study in Lima found that the level of knowledge of the disease influenced the threshold of symptoms for which people with TB decided to pursue medical attention [12]. Together, these findings suggest that community education strategies aiming to increase awareness of TB could lead to individuals seeking care earlier in the disease, and our study participants identified this as an important area for improvement. Unfortunately, previous research has not shown an impact of educational activities alone on TB case notification, time to treatment, or long-term prevalence of TB [18]. One study from Colombia found that a mass media campaign led to a sudden increase in case detection, but the effect was short-lived and not sustained [19]. Despite this, programs using a combined approach of active case finding with health education have proven successful, suggesting that this might be an important component of programs aiming to reduce TB diagnosis delays [18].

To reduce delays in people with TB getting to health facilities capable of diagnosing them, the mismatch between what the health system offers and the needs and healthcare-seeking behavior of people with TB must be addressed. The incompatibility of public health facility operating hours with people’s work schedules as well as the long wait times contributed to people delaying care or visiting pharmacies and self-medicating. Extending the schedule of public health facilities may lead to more people with TB seeking attention from these facilities, which can diagnosed and treat TB, as their first resource. In addition, training pharmacy workers to recognize and refer people with potential symptoms of TB to the appropriate services can help reduce diagnostic delays [20]. A survey of pharmacy workers in Lima found adequate knowledge of TB and a willingness to refer people with respiratory symptoms for evaluation, although only a minority reported feeling well-informed about the disease [21]. Involving these professionals in community interventions may be an important strategy to educate and screen a group of people with TB who might not otherwise be reached.

Once they entered the health system, having to go through multiple healthcare visits to reach the diagnosis of TB was a dominant theme in participants’ narratives. Studies from diverse settings have also found that repeated visits within the same level of care was the core problem leading to diagnostic delays in TB-endemic countries [9, 22,23,24]. In our study, the need for multiple healthcare visits stemmed from the lack of health system resources and the referral of patients between providers, often alternating between the public and private sectors. While in some settings seeking initial care from a private practitioner is a risk factor for delays [9], our study revealed that many participants sought care in the private sector due to long wait times and limited availability of resources in the public sector, and that private sector pulmonologists often made TB diagnoses. Thus, our findings underscore the importance for collaboration between the public and private sectors [25]. Many countries have created arrangements with centralized government-led registries, referral of patients from the private to the public sector, bidirectional support, and non-governmental organizations sometimes mediating the two [26]. While no such systems for intersectoral collaboration currently exist in Peru, our finding that having a written note from a provider was enough to expedite the process of obtaining a diagnostic test or scheduling an appointment suggests that even relatively simple and low-cost strategies might be effective in facilitating transfers of care. Additionally, creating a specimen referral system could streamline the diagnostic process. In the Peruvian public sector, an established transport network enables all facilities to send sputum samples to designated laboratories, but the private sector is not integrated into this network [27]. Enhancing the private sector's ability to send specimens to certified laboratories presents another opportunity for intervention.

Some participants in our study perceived their providers to lack appropriate knowledge about TB. In particular, individuals with extrapulmonary TB reported that providers failed to consider extrapulmonary TB and attributed their symptoms to other conditions. These observations from people with TB in our study are corroborated by studies of providers from other settings. Concerning knowledge gaps have been identified among health personnel in many countries with high TB burdens, including Peru [28,29,30]. With extrapulmonary TB specifically, medical doctors in India identified the lack of knowledge and poor training of health personnel as a barrier to diagnosis [31]. Increasing training around TB diagnosis, including extrapulmonary TB, should be a priority for providers in primary care facilities.

There are some limitations to our study. First, we defined first contact with the healthcare system as the initial visit to a health facility, and we did not ask about encounters with non-clinicians such as pharmacy workers. Thus, we were unable to quantitatively describe healthcare seeking prior to entry into the formal health system. However, with our qualitative analysis, we were able to explore how these encounters played into the overall diagnostic delay. Second, because we relied exclusively on surveys and interviews, our findings are subject to recall error. While we were able to assess consistency between survey and interview data for the participants who completed both, we did not triangulate the data by consulting medical records or by interviewing providers. Third, only one author coded the qualitative data, which could have led to bias or code drift. However, regular meetings were conducted with a second author to review and discuss the code tree and resolve any cases of code discrepancy or redundancy. Fourth, our study took place during the COVID-19 pandemic, and some of the challenges faced by people living with TB might have been brought on by the unusual strain on the healthcare system during this time. To help tease out these factors, our group investigated and previously described the unique effects of the COVID-19 pandemic in a separate study [32]. Finally, we focused exclusively on individuals who initiated TB treatment, not capturing those who sought care but did not obtain a diagnosis or start treatment. This group may face significant additional barriers to care and is important for understanding the full spectrum of barriers to TB diagnosis and treatment. Future research should attempt to reach this group for a more comprehensive analysis of obstacles to TB care.

Conclusions

In conclusion, our findings suggest that improving community awareness, extending the hours of operation of public health facilities, increasing the training of primary care providers and pharmacists, and creating a formal referral system to facilitate transfers of care between the public and private sectors could lead to more prompt TB diagnosis. Such strategies could be key in improving linkage to care and expediting the diagnosis of TB in similar settings, leading to earlier initiation of treatment and decreased transmission of the disease.