A qualitative study of patient, caregiver, doctor and nurse views of factors influencing lumbar puncture uptake in Zambia

Abstract

Background

Uptake of lumbar puncture (LP) remains low in regions with a high prevalence of central nervous system (CNS) infections like Zambia. Efforts to improve uptake are hindered by limited understanding of factors influencing LP uptake.

Methods

Semistructured qualitative interviews were conducted with patients with suspected CNS infection, caregivers, doctors and nurses at the University Teaching Hospitals in 2016. Questions focused on LP experiences, knowledge, the consent process and health system barriers to LP among patients with an LP indication. Interviews were transcribed, translated to English and analysed using a thematic approach.

Results

We recruited 24 adult patients, 36 caregivers of adult patients, 63 caregivers of paediatric patients, 20 doctors and 30 nurses (173 in total). LP barriers arose from both patients/caregivers and health providers and included community apprehension about LP, proxy (family) consensus consent practices, competing clinical demands, wariness of patient/caregiver responses, limitations in consumables and time to complete the LP. This could result in consent not being obtained correctly. LP enablers included patient/caregiver perceived LP utility, provider comfort with LP and in-person counselling.

Conclusions

LP uptake is a complex sociocultural process influenced by patient, healthcare and community-level factors. Interventions to improve uptake must address multiple barriers to be successful.

Introduction

Meningitis is the second greatest contributor to neurological disability-adjusted life years for the Eastern sub-Saharan African region.¹ In Zambia, meningitis mortality is higher than expected for its sociodemographic index, suggesting that additional factors influence outcomes.¹ Among patients with symptoms indicating central nervous system (CNS) infection, timely completion of a lumbar puncture (LP) for cerebrospinal fluid (CSF) analysis is essential to determine disease aetiology and guide treatment selection and duration.² Further, serial LPs are associated with improved survival in cases of cryptococcal meningitis where elevated CSF pressure can result in death.³ Yet, in 2011, 40% of HIV-positive Zambian adults with new-onset seizure, which could be a marker of meningitis, did not have CSF collected.⁴ Low LP uptake was also reported among Zambian children presenting with suspected neurological infection during 2016–2018; similar findings have been reported in other regions, including Malaysia and Pakistan.^5–7

Literature examining reasons for low LP completion rates has predominantly focused on patient/caregiver refusal to provide LP consent due to limited knowledge.^8–10 LP-related knowledge and attitude surveys performed in Asia and the Middle East among healthy adults or relatives of children with febrile seizures suggest that understanding of LP is limited.^9,11,12 In northern Zambia and Botswana, adults reported concerns that LP can cause death, but the relationship of this belief to LP uptake was not assessed.^13,14 Healthcare workers (HCWs) have acknowledged a heightened fear of death associated with LP in low-to-middle-income countries, possibly as a result of misattribution of high mortality from opportunistic infections.¹⁵ However, there is a lack of literature understanding this fear. Additionally, factors associated with LP uptake in paediatric patients in sub-Saharan Africa have not been evaluated.^13,14

A recent study by Saylor et al. found that LP uptake varied between three facilities in Zambia, highlighting the importance of other factors besides patient/caregiver perceptions contributing to the successful completion of an LP in a patient with a clinical indication for CSF collection (i.e. LP uptake).¹⁶ These include HCW recognition of LP indication, insufficient time to deliver information about LP and obtain informed consent, HCW perceptions surrounding LP safety and acceptance, procedural knowledge and other health system factors such as the availability of consumables needed to obtain or analyse CSF.¹⁵ The importance of these factors is evident in CSF studies being recently added to the WHO's essential in vitro diagnostics list intended to guide worldwide laboratory supply procurement.¹⁷

Unfortunately, the impact of these real and perceived barriers on LP uptake remain poorly elucidated. Few studies have examined LP barriers from the perspective of HCWs.¹⁴ Doctors in Botswana reported confusion about LP contraindications, but the impact of this on willingness to request or complete the procedure was not elucidated.¹⁴ Further, despite playing an integral role in patient and caregiver education, nurses’ perspectives have not been characterised. Attempts to improve LP uptake are likely to be ineffective until the factors contributing to this diagnostic gap are fully explored. Therefore, we conducted semistructured interviews of patients with suspected meningitis, their caregivers, doctors and nurses to characterise inpatient LP barriers.

Materials and Methods

Study design and participants

We recruited participants from the University of Zambia's University Teaching Hospitals (UTH) Adult and Children's Hospitals in Lusaka by going bed-to-bed in the adult and paediatric emergency, admission and medical wards from 4 April to 19 July 2016. Eligible patients included individuals of any age presenting with symptoms indicating CNS infection. If the patient was aged <18 y or unable to make medical decisions, a bedside caregiver was approached for participation. We opted for this approach as caregivers often serve as proxies for patient consent in this setting. Only one caregiver was interviewed per patient. Patients and caregivers were interviewed by nurses in the language of their choice. Eligible HCWs included doctors and nurses providing care on wards where patients were recruited. Nurses were interviewed by a nurse, while doctors were interviewed by a visiting physician. All interviews were conducted in a private room adjacent to the ward and, upon completion, participants were given 30 Kwacha (approx. US$6 at the time) for transportation costs. Interviews were conducted at a time convenient for the participant and interviewers were trained to recognise participant distress and medical emergencies. Interviews were halted if either of these situations arose.

Procedures

Semistructured interviews were conducted as part of a larger mixed-methods study of LP knowledge, attitudes and beliefs. Qualitative responses were hand-transcribed during the interview. Audio recordings were not conducted because of concerns they may have limited participants’ candor. Within 24 h of the interview, responses were translated to English, as needed, verified for accuracy by peer review among interviewers and entered electronically into Microsoft Excel (Redmond, Washington, USA).

Patients and caregivers

Participants were asked if they knew someone who had undergone LP. Among those that did, probes included their relationship to the individual and LP outcome. Patients and caregivers were asked how much they felt they knew about LP. Follow-up questions focused on the purpose of LP, as well as the benefits and risks of the procedure. They were asked whether patients have the right to decline LP, risks to the patient if an LP were refused and alternative procedures that could be performed if an LP was refused. Probes sought to elucidate the difference between preadmission LP knowledge and information shared by HCWs during admission. Patients/caregivers were asked to describe their concerns about LP. They were asked who they discussed the procedure with when deciding whether to consent and why they opted to consent to or decline LP. For each patient/caregiver, LP uptake was determined. This was defined as successfully performing an LP for CSF collection in a patient with an indication for the procedure.

Doctors and nurses

HCWs were asked to describe the purpose of LP, the risks and benefits and alternative procedures if an LP is refused. They were asked to detail the risks of not performing an LP and whether patients/caregivers had a right to refuse. Probes helped to clarify what information they regularly shared with patients/caregivers when discussing LP. Lastly, HCWs were asked to describe other circumstances, besides refusal, that result in poor LP uptake.

Data analysis

Analysis was conducted using a modified thematic approach. Two coauthors independently reviewed the data multiple times to understand meaning and identified significant phrases that were then restated in general terms to be used as themes. Responses were then coded and themes were organised into broader categories. Key quotes exemplifying themes were identified and provided to two additional coauthors who were asked to group the quotes into themes. Findings were then compared and discussed until consensus on the themes and categories was achieved. To determine whether themes differed between the Adult and Children's Hospitals, responses were examined by hospital. All authors agreed with the results and chose the quotations reported in the Results section.

Results

Description of study participants

Our sample consisted of 24 adult patients, 36 caregivers of adult patients, 63 caregivers of paediatric patients, 20 doctors and 30 nurses (173 participants in total). Fifty-eight participants were male and most were aged 24–48 y. Among patients and caregivers, 69 had completed at least some secondary education, 22 had formal employment whereas 38 were informally employed.

LPs were performed on 67 of 123 interviewed patients/caregivers. Seven themes emerged that could be grouped into two thematic categories: LP barriers and LP enablers. Themes were also grouped according to the actors involved (Figure 1).

Figure 1.

Factors influencing LP completion.

LP barriers

Community apprehensions

While most participants knew that an LP entailed removal of fluid from the back, there was consensus that there was a lack of knowledge about LP in the community. Despite this, most patients/caregivers knew a relative or acquaintance who had an LP and had a poor outcome. Patients/caregivers drew on these prior experiences to inform their decision whether to consent or decline LP. Death and paralysis were the most cited reported outcomes of LP:

I fear that my child might die because my father died after the procedure (paediatric caregiver).

Death was specifically attributed to timing of the procedure, if performed too late or when the patient was not strong enough. Paralysis was attributed to the position the patient was placed in during or after the procedure. Fear of death or paralysis was cited by all adult patients/caregivers and 12 paediatric caregivers as the reason for LP refusal. However, patients and caregivers felt that nowadays patients were more likely to survive LP than in the past and attributed this to a new technique for performing the procedure or, in some instances, to ‘God's will’.

Multiple physicians said that, because LP gained prominence when no treatment was available for HIV, the community associated it with the significant mortality of that era. HCWs were cognizant of the impact that community apprehensions have on patient/caregiver willingness to undergo LP. Patients may rescind consent at the urging of other patients/caregivers on the ward if too much time passes between consent and LP completion. Physicians also reported that community apprehensions influenced which patients they sought to LP. Recalling patients on whom an LP was clinically indicated but not performed due to the severity of illness on presentation, they recollected their concern that other patients and caregivers on the ward would see that patient's demise as evidence that LP kills:

If a child is in the terminal stage, then it is appropriate not to do it as it will reinforce the belief of death (paediatric physician).

Proxy family consensus for LP consent

HCWs and caregivers noted that being asked to consent on the patient's behalf resulted in psychological stress that impacted decision-making abilities:

It's hard to make a decision alone unless you consult other family members (adult caregiver).

Caregivers often consulted extended family members, often citing older male relatives or grandmothers, and the decision to consent to LP was made by consensus:

I think the decision comes from others because they don't want to take responsibility; they feel the family will blame them (adult physician).

Family consensus was a major barrier to care as consulted relatives often did not reside in Lusaka. Among those interviewed who had not yet consented to an LP, 7% of patients and caregivers were awaiting additional family discussion to consent. One paediatric caregiver refused to consent without family consensus because ‘if a big problem occurs, I will be blamed.’ When the patient was sufficiently well to make their own medical decisions, family consensus may over-rule patient wishes:

The father and uncle to the patient refused after the patient consented and doctors could not carry out the procedure (adult nurse).

While HCWs generally respected this decision-making process, it was a source of significant frustration:

They give me a headache. Sometimes you spend a long time discussing it and then people say no (paediatric physician).

Consent was most often provided verbally. Physicians did not obtain written consent as this was not the norm for procedures and they did not want to further ‘mystify’ LP. However, they formally documented refusal in the medical file. HCWs rarely considered LP refusal as an opportunity to educate or counsel patients/caregivers.

Competing clinical demands

Physicians reported insufficient time to adequately counsel patients about LP apprehensions. Those that did not counsel decision-makers modified their consent process in the hope of obtaining consent more rapidly. They recalled minimising or omitting risks during the consent process to reduce the likelihood of LP refusal:

I told them, I am going to do a lumbar puncture and it's safe (adult physician).

HCWs caring for adult and paediatric patients recalled instances where the consent process was overlooked entirely due to concerns that protracted discussions would impact patient care. This approach to not-so-informed consent often resulted in caregivers feeling as though they had little information about the procedure even after it had been completed; 16% of the patients/caregivers interviewed who had yet to consent to or decline LP indicated they were waiting for more information from HCWs. Indeed, 25% of patients/caregivers who refused LP did so because of lack of information. Multiple caregivers, more frequently of adult patients, were told to consult relatives for consent without receiving information about the utility of LP.

In addition to insufficient time to counsel patients, HCWs identified further clinical barriers to LP uptake. These included the desire to consult other physicians about the need for LP, insufficient time to perform the procedure, inability to locate consumables such as CSF tubes or an inability to locate the patient after they have moved to a different ward. A few physicians highlighted the need to obtain a CT scan before performing an LP, which would further delay the procedure. Even if an LP is completed, HCWs noted that CSF may not be obtained (‘a dry tap’) or an insufficient sample would be collected, resulting in a need to repeat the procedure.

LP enablers

Perceived utility of LP

Patients and caregivers frequently cited the need for a diagnosis to obtain the correct treatment as justification for agreeing to LP:

It helps the doctor give the right medication after seeing the results (adult patient).

Adult patients and caregivers reported this more often than paediatric caregivers. The informed consent process may have played a role as multiple HCWs in the Adult Hospital mentioned this as part of their consent process:

We tell them that a proper diagnosis will not be reached if they refuse (adult nurse).

LP was also seen to shorten the duration of admission:

One might stay long because the doctors won't know what they are treating (paediatric caregiver).

Worry about the patient's health and severity of illness were cited as reasons for agreeing to LP. Interestingly, some caregivers declined consent due to concerns that the patient was too weak and indicated that they would reconsider once the patient became stronger. Adult patients often presented after a protracted illness at home and LP was often reconsidered later as a last resort:

We consented because the patient was getting worse, so we wanted to know the cause (adult caregiver).

While caregivers may be more amenable to LP later in the admission, HCWs often viewed this as too late to impact care:

I'm never comfortable doing LP in someone who is about to die as I think it speeds it [death] up and won't change anything (adult physician).

Perception of HCW comfort with LP

Confidence in the technical capacity of the treating physician was an enabling factor for LP uptake among some patients/caregivers:

I had nothing to worry, because I knew the doctors wanted something best for me (adult patient).

Patients and caregivers often noted that complications only developed if the procedure was performed by an unexperienced clinician. HCWs shared the concern that some physicians were not sufficiently trained to perform LP. While endorsing confidence in their ability to perform LPs, some doctors reported incorrect knowledge about LP contraindications that limited their willingness to perform LP. No doctor or nurse reported dissuading a patient/caregiver from consenting to an LP, yet some were less likely to agree to an LP themselves or their relative if requested by a physician:

No, I would not do it [an LP on myself] and, no, I would not advise it because I am not comfortable with the expertise of most doctors (adult nurse).

For some, this reluctance was reinforced by concern that LP caused iatrogenic infections:

Sterility is often not observed so you can introduce infections into the sterile CSF (paediatric physician).

In-person counselling

HCWs felt they were more likely to obtain LP consent if they allowed for multiple conversations and provided a simple explanation of the procedure. Physicians would occasionally capitalise on the communal nature of the wards and use other patients as examples of individuals who had an LP without complications. Caregivers also used fellow patients for guidance:

I consulted a lady who is also nursing her sick daughter within the ward who advised me that, even her, she underwent lumbar puncture, and nothing happened to her (paediatric caregiver).

All participant groups cited a need for further education about LP and frequently referred to in-person HIV sensitisation efforts as a well-developed educational intervention. However, in-person counselling alone was not enough to ensure LP uptake:

We were given all of the information but, as a family, we preferred blood tests before LP (adult caregiver).

Discussion

LP uptake is a complex sociocultural process involving patients, their families and community and the healthcare system and its workers. By interviewing 173 patients, caregivers, doctors and nurses, we identified multiple themes that influenced LP uptake. These included apprehension about death and paralysis, the need for proxy family consensus and wariness of HCWs in time-limited and resource-challenged settings to obtain consent, deliver counselling and perform LP correctly. However, there was an indication that LP attitudes were now more positive amongst patients/caregivers, particularly if prior LP experiences were positive, HCWs were perceived as experienced and patients/caregivers were counselled.

Perceptions of a heightened risk of death and paralysis were common among patients/caregivers and HCWs were wary of reinforcing this association by performing LP when patients were extremely ill. Concerns about LP-related risks have been commonly reported in our patient population,¹⁸ as well as elsewhere in Africa, Asia and the Middle East, and have been associated with increased LP refusal by patients and their caregivers.^8–10,19 In this study, these attitudes were a result of prior experiences with LP among seriously ill individuals as well as a limited understanding of the procedure.

The decision to consent to LP requires patients/caregivers to weigh the perceived risks of death and paralysis against possible benefits, which were frequently unclear to participants. Health literacy in Zambia is limited²⁰ and, while our data suggest that LP-related fears can be overcome as part of the informed consent process, HCWs may view consent as a barrier to time-sensitive care, particularly in the context of limited resources. Physicians often limited education about LP to expedite the consent process. Unfortunately, lack of education was cited as a reason for refusal and dissatisfaction among patients and caregivers. Task-shifting LP education and consent from physicians and nurses to dedicated patient educators may be one way to address education needs and overcome community misconceptions. An intervention such as a bedside video could be one way to improve LP education among patients and proxy decision-makers.²¹ Temsah et al. found that video-based counselling similarly improved LP-related knowledge among parents of children obtaining outpatient clinical care compared with routine in-person counselling.²² However, video-based counselling was also associated with an increased perception of LP-related risks than routine in-person counselling.²² This may inadvertently increase LP refusal, although this was not assessed.

Adding to the burden associated with the LP-consent process is the need to obtain consensus among multiple proxies for consent. In Zambia, there is not a legal precedent for proxy medical decisions and, as a result, HCWs must quickly identify decision-makers before the consent process can begin. This can be particularly challenging when a family's preferred proxies cannot be reached by telephone or cannot travel to provide consent. Family and community involvement in informed consent for clinical research has been described in multiple African and Asian settings.^23–25 As in Nigeria, where this approach is seen as a safety net for community members, adopting consent practices that expedite biomedical care at the expense of proxy involvement may further isolate patients and caregivers during a highly stressful time.²⁶ Calls to standardise informed consent practices in this setting should consider the role of family and community in decision-making.^14,26 Further, near universal multi-proxy family consent among our study population suggests that an intervention targeting just the bedside caregiver may have limited impact on increasing LP uptake as that caregiver may lack the authority to make decisions.

This study identified health system factors that contribute to poor LP uptake in our setting, including difficulty obtaining the appropriate consumables for the procedure and tracking patients if they are relocated during admission. These logistical barriers hinder LP uptake among patients who have already consented to the procedure and, until they are addressed, may deter physicians from proposing LPs on subsequent patients. If this is the case, physicians may be inadvertently reinforcing the belief that LPs are not necessary to diagnose meningitis and, therefore, limiting patient and caregiver demand for this procedure. Additional health system barriers to LP uptake could include lack of laboratory reagents for the appropriate investigations and physician confidence in CSF results, although these were not spontaneously raised by HCWs and should be examined in future studies.

Adult patients and their caregivers also reported presenting late to biomedical care for meningitis symptoms. As one of the drivers behind LP refusal was concern that the patient was too weak for the procedure, earlier referral for biomedical care may improve LP uptake as well as outcomes. Future studies should examine patient and caregiver understanding of meningitis as well as care sought in the community before up-referral to a tertiary institution.

This study was conducted at the nation's largest tertiary care centre in an inpatient setting. The barriers to LP uptake may differ in rural facilities where the access to neurological expertise is often more limited. However, HCWs in this setting may be more familiar to patients and their families and, as a result, may be better positioned to facilitate multi-proxy consent to improve LP uptake. Further, there may be additional barriers to LP in an outpatient setting where factors such as access to a procedure room, post-LP care and laboratory diagnostics may be a concern.

Conclusions

While patient and caregiver refusal to provide consent is often cited as the most salient barrier to LP completion, barriers to LP uptake are complex and reinforced by patients/caregivers and HCWs, in a cycle of community apprehensions and family consensus influencing HCWs and making them wary of obtaining consent correctly and of performing LP themselves. This is further exacerbated by limited resources and infection control issues. However, there was a shift in patient/caregiver apprehensions with usefulness, medical expertise and counselling countering concerns. Interventions to improve LP uptake in low-resource settings must be multi-faceted, include patients, caregivers and HCWs, target barriers at all levels and build on enablers.

Authors’ contributions

MAE and MPK conceived the study; MAE, GLB and MPK designed the study protocol; MAE and CB carried out data collection; MAE, CB, MPK and IS carried out analysis. All authors assisted with interpretation of these data. MAE and GLB drafted the manuscript; all authors critically revised the manuscript for intellectual content. All authors read and approved the final manuscript. MAE and MPK are the guarantors of the paper.

Funding

This work was supported by the National Institute of Neurological Disorders and Stroke [R01NS094037 {to GLB and IS}, 5R25NS065729-11 {to MAE}); and Michigan State University's Norman Kagan Endowed Scholarship in Graduate and Professional Studies and the Blake W. H. Smith Scholarship.

Competing interests

There are no conflicts of interest.

Ethical approval

Written, informed consent was obtained from participants in the language of their choice (English, Nyanja or Bemba). Ethical approval was obtained from the University of Zambia's Biomedical Research Ethics Committee and Michigan State University Biomedical Institutional Review Board.

Data availability

The data underlying this article will be shared upon reasonable request to the corresponding author.