br Healthcare pathway evaluations In our example we will mak

Healthcare pathway evaluations In our example, we will make these evaluations as assessments against hypothetical “ideal” behaviours. Our illustrative definition of “ideal” behaviour involves access to regulated healthcare providers and minimal delays to care where the illness requires the patient to do so. The assessment criteria of what counts as “ideal” are externally imposed on grounds of health system efficiency and concerns about delays in treatment (explained below and exemplified in Fig. 8). This does not necessarily mean that patients or healthcare providers adhere to the same criteria when they purchase KN-93 make their choices, and their personal notions of an “ideal” process can very plausibly differ (and for good reasons, including e.g. health knowledge, societal perceptions of the health condition, concerns about provider competence, or variations in decision-making autonomy; see previously cited references on healthcare-seeking determinants). Therefore, we do not judge whether a patient is right or wrong in their personal choices, but we make a judgement as to whether patient behaviour aligns with the externally imposed evaluative criteria of the analyst, which may pertain e.g. to health system efficiency. We make our evaluations with reference to public healthcare access in rural Gansu and Rajasthan, considering that our field sites are particularly resource-constrained (people in remote villages often have to walk for half an hour to reach a clinic), and the health system in rural Rajasthan is particularly fragmented with a broad range of limitedly regulated private and informal healthcare providers. We exemplify the application of sequence-sensitive evaluations by examining if wealthier households are more likely behave in line with (illustrative) “ideal” processes of healthcare utilisation, immediacy of treatment from public healthcare, and adherence to referral procedures. Other researchers may want to articulate specific hypotheses about the determinants of healthcare-seeking, examine behaviours for different health issues with fundamentally different pathway profiles (e.g. communicable vs. non-communicable diseases, childbirth), or make different assumptions as to what constitutes ideal healthcare-seeking processes depending on their study focus and the local health system context. Our methods allow such a flexible and transparent adaptation of the analysis. Fig. 8 exemplifies our three illustrative approaches to evaluating “ideal” healthcare pathways, assuming that a patient should get treatment from a public healthcare provider (see below on how we judge whether medical care is required). Panel a displays hypothetical “ideal” pathways to public healthcare; Panel b contains examples where such an “ideal” pathway is not adhered to. The evaluation of hypothetical “Ideal a” follows the conventional, sequence-insensitive approach and considers only whether any public healthcare access took place during an illness, leading to a negative evaluation where this is not the case (see Panel b). The evaluation of “Ideal b” considers the process of accessing public providers, with a negative evaluation if referral systems are bypassed. Whereas the “ideal” process in Panel a involved a visit to a local doctor prior to public hospital access, the “ideal” is violated if did not contain such a first-tier provider visit. Lastly, from a health system utilisation perspective, analysts might also be interested whether patients delayed their treatment by approaching untrained, unregulated, or costly private and informal practitioners prior to visiting public healthcare providers. The evaluation of “Ideal c” in Panel a demonstrates accordingly how a healthcare pathway might comply with this “ideal” if no such contact took place before the public doctors and hospitals were accessed, in contrast to the violation of the hypothetical “ideal” in Panel b. Such process benchmarks are clearly normative and context-specific, but they can be tailored and specified transparently if sequential healthcare-seeking data is available, thus going beyond binary measures of “access” or “no access.”