The South African National Policy Framework and Strategy on Palliative Care (NPFSPC) recommends that when integrating palliative care (PC) into the health system, a PC indicators tool should be used to guide clinicians to recognise a patient who should receive PC. The policy document recommends ‘a simple screening tool developed for use in South Africa that would assist healthcare professionals (HCPs) to recognise patients who may have unmet palliative care needs’.
This research study sought to develop South African consensus on indicators for PC to assist clinicians to recognise a patient in need of PC.
The South African healthcare setting.
A Delphi study was considered suitable as a methodology to develop consensus. The methodology was based on the Conducting and REporting of DElphi studies (CREDES) guidance on Delphi studies to ensure rigour and transparency in conducting and reporting. Six different Delphi rounds were used to develop consensus. Each round allowed participants to anonymously rate statements with predefined rating scales.
Cognisant of the disparities in healthcare provision and access to equitable healthcare in South Africa, the expert advisory group recommended, especially for South Africa, that ‘this tool is for deteriorating patients with an advanced life-limiting illness where all available and appropriate management for underlying illnesses and reversible complications has been offered’. The expert advisory group felt that disease-specific indicators should be described before the general indicators in the South African indicators tool, so all users of the tool orientate themselves to the disease categories first. This study included three new domains to address the South African context: trauma, infectious diseases and haematological diseases. General indicators for PC aligned with the original Supportive and Palliative Care Indicators Tool (SPICT) tool.
The Supportive and Palliative Care Indicators Tool for South Africa (SPICTTM-SA) is a simple screening tool for South Africa that may assist HCPs to recognise patients who may have unmet PC needs.
The World Health Assembly 67.19 resolution on palliative care (PC) states that the provision of PC is ‘an ethical responsibility of health systems’, and it is the ‘duty of healthcare professionals (HCPs) to alleviate pain and suffering’.
The NPFSPC policy document recommends ‘a simple screening tool developed for use in South Africa that would assist HCP to recognise patients who may have unmet PC needs’ and identified the Supportive and Palliative Care Indicators Tool (SPICTTM) and the Gold Standards Framework Prognostic Indicator Guidance (GSF-PIG) as two such validated tools, which comprise disease-specific indicators and general indicators of a progressive illness and help to identify a need for PC.
The World Health Organization definition of PC clearly states that PC ‘is applicable early in the course of illness, in conjunction with other therapies that are intended to prolong life’.
The utility and validity of PC indicator tools developed and used in other parts of the world may be limited for the South African context because of differences in the socio-economic environments and the availability of resources for healthcare. South Africa has extreme socio-economic inequality, evidenced by the high Gini coefficient.
This research study aimed to develop the South African consensus on indicators for PC to assist clinicians to recognise a patient in need of PC. The objectives of the study were to identify panellists who were representative of the South African healthcare context, to ask them to respond to a survey of possible indicators for PC, to come to a consensus through a Delphi process and to have an external group review the consensus document
A Delphi study (
Flow diagram of the study methodology using the Delphi approach.
By applying the CREDE tool, bandwagoning was prevented, and anonymity was ensured by allowing the expert group to provide individual input and by using an anonymous online survey tool (Research Electronic Data Capture [REDCAP]) for the panellist. Rating scales were predefined and the definition of consensus was predefined for all the Delphi rounds. Consensus was defined when ≥ 80% of the participants rated a statement as ‘agreed’ or ‘strongly agreed’. Statements with ‘moderate’ to ‘low agreement’ were fed to the next Delphi round. If no consensus was reached and arbitrator from outside would be appointed.
A research team consisting of three PC practitioners with more than 15 years of experience in the field convened an expert group. The expert advisory group was formed consisting of South African healthcare workers who were deemed experts because they had more than five years of experience in their field, they were trained in PC and regularly referred to PC. This group included three PC clinicians, one general surgeon, two specialist physicians, two specialist haematologists, one social worker, two oncologists, two family physicians, one emergency medicine specialist, one registered nurse and one specialist geriatrician.
Ethics approval was received from University of Cape Town Human Research Ethics Committee (HREC 750/2018 and R012/2016).
The expert advisory group met face-to-face to reach a consensus on the role of a PC indicators tool for South Africa. This consensus was intended to meet the requirements of the first step in the Delphi process (Delphi 1A).
The initial consensus in January 2019 on the purpose of the tool was that:
The indicators tool aims to identify adult patients who have serious health-related suffering due to life-limiting or life-threatening illness and whose health is deteriorating. Patients who suffer from such conditions are likely to die as a consequence. This is a generic tool for the South African setting to identify patients who will benefit from a palliative care approach in conjunction with usual care by the treating clinician. This patient may need to be referred to a palliative care team to optimise care.
The expert advisory group members were provided with reference materials describing international indicator tools. They developed a draft survey tool for the first Delphi round by specifying domains of illness and indicators for PC based on their field of expertise. All members also submitted general indicators of deteriorating health, which would indicate a need for PC. In order to preserve independence of contributions, members recorded their thoughts individually and were encouraged not to share these during the initial process. The research team analysed all the submitted indicators, deleted duplications and collated all indicators into a survey document. This was circulated to the expert advisory group in October 2019 for final comment.
All expert group members were asked to identify colleagues from across South Africa who met the inclusion criteria as panellists for the second phase of the process, Delphi 2. These were (1) South African HCP, including specialist doctors, general practitioners, registered nurses, allied healthcare professionals and social workers who have been trained in PC or who regularly refer patients to PC services, (2) having more than five years’ experience in their relevant speciality or discipline and (3) considered by the expert advisory group to be expert in their field. This tool aims to develop an indicators tool for adults and thus paediatricians were excluded.
The final survey document was sent out via the online anonymous survey tool called REDCAP, a secure web platform for building and managing online databases and surveys, to 116 professionals across South Africa in November 2019. The professions of the panellists are listed in
Panellists for Delphi 2.
Profession | Number |
---|---|
Family physicians | 11 |
Geriatricians | 13 |
Emergency medicine physicians | 1 |
Internal medicine physicians | 28 |
Surgeons | 16 |
Gynaecologists | 3 |
Haematologists | 7 |
Registered nurses | 12 |
Social workers | 2 |
PC trained physicians | 11 |
PC, palliative care.
The response rate to the second phase of the process (Delphi 2) was 37% and came from 43 participants across South Africa. Two reminders were sent via email although these had a limited effect on the response rate.
Some participants contributed statements out of keeping with common practice in PC, but these were retained to ensure complete capture of all ideas concerning PC from the surveyed sample. This raised a concern that even experienced HCP may not grasp the role of PC. This further supported the investigators’ intention of providing a tool to identify patients who would benefit from PC, in South Africa.
The responses of the 43 participants were collated and reviewed at a second meeting of the expert advisory group, convened in January 2020. A domain was included if the majority (median ≥ 80%) of the panel had scored it at 8 or more out of a maximum of 10 (1/10 being not for inclusion and 10/10 being for definite inclusion). The domains agreed for inclusion were respiratory, renal, hepatic, neurological, cardiovascular, haematological and infectious diseases and cancer and dementia. The expert advisory group recommended that frailty be added, as in the original SPICTTM, as a distinct element alongside dementia.
Domains where agreement was not reached were trauma, mental health and gastrointestinal disease. Most participants scored these at 6–8/10. Trauma, infectious disease and haematological disease were included in the possible list for inclusion based on South African mortality and morbidity data.
Even though the general indicators of increasing dependence as formulated by the expert advisory group showed consensus, it was felt that these were too loosely described to be useful. The expert advisory group felt that the tool should however aim to include four to five general indicators. There was consensus by the expert advisory group after another online survey that the developers of the original SPICTTM had already performed a very comprehensive review of these general indicators (
Number of patients with disease-specific indicators referred to a palliative care service.
A new draft indicators tool was created taking into account consensus statements and discussions with the expert advisory group. This draft was sent to the 116 panellists via an email link to the REDCap survey in February 2020. There were only 14 responses in this round and no changes were made. This low response rate does limit the findings of this study and further clinical validation will be required to strengthen the validity of this tool. This will be achieved by applying the tool on a patient population and alongside evaluation to determine whether the patients have unmet PC needs.
The new draft indicators tool was externally reviewed by the SPICT programme lead in Edinburgh in Scotland. The Edinburgh group guided the research team, drawing from previous experience in developing and using indicators tools and best practice in adaptation of clinical tools for different settings. This included use of widely accepted language and concepts and the process of consensus building. The final recommended changes were circulated to the expert advisory group for comments after the Edinburgh review.
The original SPICTTM research team has used participatory research to evaluate their tool in clinical practice and concluded that the tool can support clinical judgement by multidisciplinary teams when identifying patients at risk of deteriorating and dying.
The final SPICTTM-SA (
The supportive and palliative care indicators tool South Africa tool.
Implementing the World Health Assembly (WHA) resolution recommendation that all HCPs and health workers should be trained in PC means that all healthcare professionals should integrate PC into the care of patients with life-threatening diagnoses.
A PC Indicators Tool is a helpful tool to guide healthcare workers to recognise disease specific and general indicators for PC. In developing this SPICTTM-SA, there was consensus that general indicators of deteriorating health remain universal across healthcare settings. This tool differs from the original SPICT tool by placing disease-specific indicators before general indicators. Supportive and PC indicators tool South Africa also includes more disease-specific indicators, namely trauma, haematological diseases and infectious diseases.
Contextualisation of disease-specific indicators is required to address the specific disease profile of South Africa and to adapt to the availability of different resources in South Africa. We recognise the limitations of this study in particular the fall off of feedback and input in the later stages of the Delphi process. The low response rate (12%) in Delphi 4 may call into question the validity of the tool. This will be tested during the planned validation study.
Palliative care must never be a substitute for appropriate disease-directed care but rather as supportive therapy that should be offered alongside the best available treatment.
The SPICTTM-SA requires validation in the South African and comparable African settings to ensure that it is robust in practical applications and that it accurately identifies patients who would benefit from early PC to reduce health-related suffering because of serious illness. The SPICTTM-SA already demonstrates the specific diseases referred to PC services as indicated in the referral pattern to a PC service in a tertiary hospital in South Africa (
The SPICTTM-SA is a simple screening tool for South Africa that may assist HCP to recognise patients who may have unmet PC needs. In addition to previously described diseases, this tool includes infectious diseases, trauma and haematologic diseases, which reflects the current South African disease burden.
The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.
R.K., A.B., H.B., A.d.V., K.E., L.F., N.F., S.K., J.M., Z.M., E.P., T.R., P.R., E.V., K.B. and L.G. contributed equally to the design of this article. R.K. and L.G. completed the analysis and the writing of this article.
The researcher (R.K.) received funding for PhD research from the Harry Crosley fund.
The data are available on an encoded RedCap database. Access can be obtained with University of Cape Town Health Research Ethics Committee permission (e-mail:
The views and authorship expressed in this article are of the authors and do not necessarily reflect the official policy or position of their institutions.