Analysis of early diagnostic pathway for prostate cancer in Slovenia

Prostate cancer is a common malignant tumour and has the highest incidence of all non-cutaneous cancers worldwide in males. It is also the fifth leading cause of cancer death among men in 2020.¹ In Slovenia the 2020 age-standardized (World standard population) incidence rate was 62.7.² In Central and Eastern Europe it was 46.4 and 59.1 in Southern Europe.¹ The ageing population undoubtedly influences the increasing incidence of prostate cancer. However, the proportional growth of elderly individuals (aged 65 years and older) in Slovenia has not escalated sufficiently to solely account for this rise in prostate cancer incidence. For example, in 1991, the proportion of elderly people in Slovenia was 11.2% and the crude incidence rate for prostate cancer was 26.3 per 100,000 individuals. In 2001, the proportion of elderly increased to 14.3% and the crude incidence rate for prostate cancer was 74.5.^3,4 The dramatic increase in the incidence of prostate cancer over the past two decades is not due to any newly identified risk factor, but rather to the increasing use of the prostate-specific antigen (PSA) test in healthy men and therefore detection of a large number of cancers that would have otherwise remained undetected for life. Data for recent years indicate that we have already reached the peak incidence of prostate cancer.² The histological incidence far surpasses the proportion of individuals in whom the disease is manifested, and while there has always been a tendency not to misdiagnose those with prostate cancer, there is increasing focus on identifying patients who are treated by watchful waiting. Prostate cancer is characterised by a slow natural course of the disease, with the majority of patients dying from other, non-cancer related causes.^5,6 Both the incidence and mortality rates of prostate cancer in Slovenia are above the European average. In 2020, Slovenia reported an age-standardized (World standard population) mortality rate of 14.9, while Central and Eastern Europe recorded a rate of 13.7, the highest among all European regions.^1,2

Much effort has already been applied to optimize early detection in prostate cancer and some medical professional societies and organizations recommend prostate cancer screening or discussing screenings with men of suitable age and life expectancy.⁷ Many countries in Europe have joined PRAISE-U project to establish screening programme for prostate cancer.⁸ In primary care-based health care systems, in which the general practitioner (GP) is the patient's first contact and triages the patient's further access to the system, most prostate cancer patients either present to a GP with symptoms (dysuria, lower urinary tract symptoms (LUTS), abdominal pain) or an elevated level of PSA is found in asymptomatic patients. Therefore, timely recognition of cancer-related complaints and adequate referral by the GP are and will remain essential to reduce time to diagnosis until successful screening programmes are introduced. Even though the association between time intervals in the diagnostic pathway and clinical outcomes is complex and remains debated, evidence suggests an association between shorter times to diagnosis and more favourable outcomes in breast cancer, colorectal, head and neck, testicular cancer and melanoma.^9,10 Optimising the diagnostic pathway from first presentation to diagnosis and start of treatment, usually interpreted as shortening the diagnostic phase, has therefore been a main objective of health care organisations involved in cancer care worldwide. For some countries in Europe, the duration of several of these intervals has been charted.^{11,12,13,14,15,16,17} For other countries, such as Slovenia, the duration of these intervals is unknown. International comparison of the duration of diagnostic intervals in different health care systems and cultural environments is important to identify system-, disease- and patient-related factors that contribute to an unnecessarily prolonged patient journey. Exploring the duration of the diagnostic pathway in Slovenia and how primary care contributes to it generates relevant information on international differences in the duration of the diagnostic pathway. This provides the opportunity to distinguish underlying mechanisms of delay, including system-, disease- and patient-related delay.

The aims of the study were (i) to evaluate the duration and specifics of management during the doctor and primary care intervals in Slovenia, (ii) to investigate the potential association between the durations of the intervals, sociodemographic and organizational variables and (iii) to assess the association between these variables, intervals and the 5-year survival rates of prostate cancer patients in Slovenia.

This study provides an overview of early management of patients with prostate cancer in family medicine in Slovenia. Our aim was to evaluate the treatment timelines and specificities within the doctor and primary care interval. The median length of the doctor interval was very short at 0 days, the primary care interval exhibited a median duration of 5 days. Our study demonstrated statistically significant correlation of cancer stage, comorbidities, and the accessibility of laboratory tests within primary health care centres and survival of prostate cancer patients. The presence of laboratory tests within primary health care centres emerged as a significant determinant of the primary care interval's duration and 5-year survival.

The exclusion of 617 patients' of 1431 patients total from our analysis was necessitated by various reasons outlined in Figure 1. It is worth noting that the proportion of cases excluded due to missing interval data was similar in our study to previous studies.¹¹

Evidence about the length of the doctor interval in patients with different cancers is sparse. We found a study of Denmark's cancer patients²⁶, where median duration of doctor interval for prostate cancer patients was 0 days and interquartile range (IQR) 0–6 days, which is the same as in our study. One of the reasons for a short doctor interval is that in primary-level diagnostic procedures, doctors sometimes use laboratory and ultrasound tests already performed prior to this first visit (e.g. blood and urine taken from the patient a month earlier, or an abdominal ultrasound performed three months earlier - doctor's interval is negative, marked as 0 days in the analysis). In this way they omit unnecessary duplication of tests and expedite further management of suspected prostate cancer.

There were no statistical differences in the duration of doctor interval between different age groups, levels of education, region of primary health care centre and accessibility to abdominal ultrasound in the primary health care centre. The duration was statistically significant longer when patients had no comorbidities, when they first presented with symptoms in comparison to elevated PSA value and where there was not access to laboratory diagnostics (CBC, CRP, urine, PSA) in primary health care centre. We interpret these results by assuming that multimorbid patients see their doctor more often, for different reasons, and thus are more likely to report or be asked about different symptoms and signs, as opposed to patients without comorbidities. In patients with elevated PSA levels in the blood, the guidelines recommend that prostate cancer should be excluded, so the decision to refer to a urologist was easy and quick. Dysuria, LUTS and abdominal pain are typical symptoms of a number of different diseases and are not in themselves of great predictive value for a cancer diagnosis.^27,28,29 Men experiencing urinary problems were more inclined to seek medical attention compared to asymptomatic men, leading to more frequent PSA testing and consequently earlier detection of prostate cancer. Our study focused on diagnosing prostate cancer cases in 2014 rather than estimating the overall prevalence that year. Therefore, it's probable that symptomatic cases outnumber asymptomatic ones. In Slovenian primary health care centers, PSA testing is more commonly conducted for symptomatic men or as part of annual check-ups for those undergoing therapy for benign prostatic hyperplasia, rather than for routine screening of asymptomatic individuals. Interestingly, we found that patients with stage spread cancer had the shortest doctor interval, while patients with metastatic cancer had the longest. This may be due to non-specific or absent symptoms and signs in localised prostate cancer. Symptoms and signs become more frequent and pronounced in spread stage of cancer and those caused by metastatic cancer could also be wrongly attributed to other patient's comorbidities. Considering health system factors, international comparisons suggests that the problem of early cancer diagnosis is ubiquitous across contemporary health systems, including high-income countries, though the same underlying problem is manifested differently depending on health service organisation, healthcare professional cultures, and the public understanding of cancer.¹⁰ This is the first study using Slovenian data to show a correlation between not having easy access to laboratory diagnostics (CBC, CRP, urine, PSA) in primary health care centres, longer duration of doctor and primary care interval and worse 5-year survival. Considering the impact of the length of the intervals on 5-year survival, we assume that the longer intervals are indicative of the broader health care situation (the influence of the patient, the doctor and the local health care system characteristics), which as a whole influenced the survival of this patient.

The duration of the primary care interval has been investigated in several studies.^{11,13,14,17,30,31} The median duration of the primary care interval in our study was 5 days, with an interquartile range (IQR) extending from 5 to 58 days which is shorter than in Helsper et al study, where the median was 14 days (IQR 3-153).¹⁵ Lyratzopoulos et al¹¹ reported the mean length of primary care interval in prostate cancer patients 31 days (25^th Centile 2 days, 90^th Centile 74 days) whereas in our study the mean was 67.4 days with SD 123.8. The duration was, similar to doctor interval, statistically significant longer when patients had no comorbidities, when they first presented with symptoms in comparison to elevated PSA value and where there was no direct access to laboratory diagnostics (CBC, CRP, urine, PSA) in primary health care centres. Surprisingly, we observed a significant difference between the Osrednjeslovenska and Savinjska regions, with the latter showing a shorter primary care interval. This was the only parameter that showed difference in the Slovenian regions. Osrednjeslovenska region is the most densely populated, the largest in terms of number of inhabitants and the second largest in terms of area. In 2014 the Savinjska region had less than half as many inhabitants as the Osrednjeslovenska region and had the lowest proportion of the population aged over 80 years in the country (4.3%). Osrednjeslovenska region had the highest gross domestic product (GDP) per capita in the country (25.329 €/resident), 11.7% registered unemployment rate and 11.4% of people at risk of poverty. Savinjska region had a regional GDP of 16.455 €/resident, 13.9% registered unemployment rate and 15% of people at risk of poverty.³² Lower socio-economic factors are usually associated with longer intervals.^12,33,34 In 2014 Savinjska and Osrednjeslovenska region had the same age-standardized prostate cancer incidence rate (99.2 vs. 100.1). Savinjska region had a higher standardized death rate (574.8 vs. 463.0) and age-standardized death rate due to neoplasms (231.4 vs. 189.4) than Osrednjeslovenska region.³⁵ Taking all these data into account, it is difficult to explain the shorter primary care interval in the Savinjska region.

In this study we also aimed to explore potential associations between the duration of primary care interval and the 5-year survival rates among prostate cancer patients in Slovenia. As expected we found worse 5-year survival in older patients, with higher cancer stages, lower education and more comorbidities.²⁵ Survival was not influenced by the region where the GP worked. We found a statistically significant difference in 5-year survival in patients with lower access to laboratory tests and abdominal ultrasound by their GP, and in symptomatic patients compared to those with a first contact due to a detected elevated PSA level. Other research has also confirmed that the quality and speed of the diagnostic process is influenced by a number of factors beyond the diagnostic skills of individual clinicians, such as the tests and services available to them, time constraints to consultation duration and the quality of doctor-patient communication.¹⁰

The use of the Slovenian Cancer Registry database enabled a comprehensive analysis as a substantial number of patients diagnosed with prostate cancer in 2014 were included. This large cohort provides robust statistical power and improves the generalizability of the study results. By using registry information, all patients diagnosed with prostate cancer in 2014 were identified and included in the study, minimizing potential selection bias and ensuring a representative sample. The limitation was the 58% response rate. As part of the study, GPs were encouraged to review both paper and electronic patient records, including discharge letters, to ensure comprehensive and accurate reporting. This approach aimed to reduce recall bias and increase the validity of the results.

The retrospective nature of the study, relying on established diagnoses and historical data, brings inherent limitations. In particular, the study focused on the doctor and primary care interval rather than the patient interval, which could affect the accuracy of reported symptoms and timing. Despite efforts to minimize recall errors, incomplete information in some questionnaires may have led to information bias, particularly with regard to the duration of doctor and primary care intervals. This bias could underestimate the actual intervals as it is based on fragment-ed or incomplete patient records. The change of GP chosen by patients between diagnosis and data collection, coupled with the variability of medical record systems, posed a challenge in accessing complete and consistent data. Paper records were frequently relied upon, which were often incomplete and difficult to decipher, while electronic records experienced compatibility issues, further complicating data retrieval and potentially leading to missing or inaccurate information. The presence of missing data leads to uncertainty about the actual performance or recording of diagnostic procedures, such as digital rectal examinations, which could affect the completeness and reliability of results. The same applies to shared decision-making. Clinical practice guidelines on prostate cancer screening using the PSA test have clearly recommended that clinicians practice shared decision making - a process involving clinician-patient discussion of the pros, cons, and uncertainties of screening. However, studies have shown that most men have never engaged in shared decision-making conversations with a healthcare provider about PSA testing.^36,37 In our study, shared decision-making was not recorded in the GP's medical records, but we cannot say with certainty that it was not carried out either.

This study provides a comprehensive examination of early management of prostate cancer patients within the primary healthcare system in Slovenia and offers valuable insights into diagnostic timelines and their impact on patient outcomes. The significant correlations found between the duration of intervals, 5-year survival rates and accessibility of laboratory diagnostics emphasize the crucial role of timely and comprehensive diagnostic testing in improving prognosis. From a clinical perspective, these findings underscore the importance of ensuring adequate resources and infrastructure for diagnostic testing in primary health care centres to minimize delays in diagnosis and optimize patient care. In addition, the observed regional differences in diagnostic intervals highlight the need for targeted interventions to address inequalities in access to healthcare and improve diagnostic efficiency in different regions. From a research perspective, these findings provide a basis for further investigation of the underlying factors influencing diagnostic processes and the effectiveness of interventions aimed at reducing diagnostic delays. Future research efforts should prioritize the validation of these findings through longitudinal studies and comparative analyses, while exploring new strategies to increase diagnostic efficiency and improve patient outcomes in prostate cancer. This study also emphasizes the need for further investigation of the factors that influence an individual's decision to seek primary health care services, particularly those beyond the scope of the variables included in the current model.