Real-life data from an oncology infirmary in a Brazilian universitybased teaching hospital

Oncological patients are mostly treated in ambulatory scenarios but may demand in-hospital care and even support at intensive care units (ICU). Little data has been published on the mortality statistics of an oncological cancer infirmary with most studies so far being developed in ICU scenarios.(1) To our knowledge, this information is a key for a better understanding of patient's needs and development of clinical assistance. Our university-teaching hospital is a reference cancer center for metropolitan area of Sorocaba (State of Sao Paulo, Brazil), with 27 nearby cities total population: 2.120.095 habitants, according to official government estimate. The service conducts approximately 15,000 ambulatory consults yearly. After a thorough literature search, we did not find any papers analyzing this matter, reinforcing the importance of this work.

Analyze data regarding patients in need of hospital care admitted in our oncology infirmary aiming better knowledge of clinical support and a real scenario for future financial and structural and professional needs for oncologic institutions.

We have designed a prospective study of 190 patients admitted for in-hospital care between 02/02/2018 and 02/28/2019, who were stratified by diagnosis, age, sex, admission reason, duration of stay, re-admission rate, and outcomes. All patients were evaluated, without loss of data. Data was show as absolute and relative numbers.

The one-year-period revealed that 97 male and 93 female patients were admitted, with a 61-year-old mean age. Based on ambulatory consults we pointed an admission rate of 1,2%. The most common primary tumor sites were: colorectal=30 (15,7%), breast=28 (14,6%), stomach=20 (10,4%), lung=14 (7,3%), ovary=12 (6,2%), and prostate=10 (5,2%). These represented more than half of the cases (59,6%), as shown inTable 1 . The mean hospital stay was 13,4 days with a re-admission rate of 9,4%.

The main admission motives were: terminality [defined by PS-ECOG=4, with a 30 day or less estimated life expectancy]: 30 patients (15,7%); total pain and or overwhelming pain: 23 (12,0%); oncological emergencies [febrile neutropenia, malignant hypercalcemia, medullar compression, superior vena cava syndrome]: 20 (10,4%); clinical deterioration [defined as worsening of patients performance or malnutrition]: 15 (7,8%); abdominal pain [not acute abdomen syndrome]: 14 (7,3%). (Table 2).

Patients outcomes in this period: 99 deaths (mortality rate 52,1%). Patients in terminality presented 93,3% ratio of mortality within an average of 6,6 days of hospital care. Febrile neutropenia (42,8%) and abdominal pain (35,7%) both achieved high mortality rates.

Among all the admissions evaluated, 91 patients had improvement or clinical stability compatible with ambulatory care and were discharged from hospital (Table 3).

Providing in-hospital care for oncological patients remains a challenge. In our analysis the in-hospital admission rate of 1,2% points at good quality of clinical ambulatory practices. This should involve not only a welltrained staff but also a comprehensive multidisciplinary care for nutritional and psychological support to these patients, not mentioning an appropriate infrastructure.

Our clinical guidelines are well based on Brazilian and international guidelines such as National Comprehensive

Cancer Network (NCCN), European Society for Medical Oncology (ESMO), Multinational Association of Supportive Care in Cancer (MASCC), and Academia Nacional de Cuidados Paliativos (ANCP). Even though our service does not have a specific palliative-care consultation team, all patients were evaluated and cared for by a well experienced staff of clinical oncologists, not mentioning a multidisciplinary team of nutritionists, physiotherapists, and psychologists. In our analysis, all patients had effective management comparisons for Brazilian services. Analysis of this group revealed that over half the individuals (N=8) were stage four (IV) metastatic solid tumor patients, namely: gastric, breast, ovary, colon, and sarcoma (Table 4).

“Malignant” hypercalcemia occurred in four patients within the studied period, affecting three female patients with breast cancer, and one male with head and neck cancer. Nonetheless, our mortality rate was at 25% while literature points at 50% death rate within 30 days of diagnosis.(5 6) The single death occurred in a triple-positive breast cancer patient who was also diagnosed with pulmonary embolism during hospital stay and presented rapid clinical deterioration despite all the support. This low incidence of hypercalcemia can be attributed for better controlled risk of hypercalcemia in ambulatory patients.

Both spinal cord compression and superior vena cava syndrome accounted for only one hospital admission each and both patients were eventually discharged from hospital care after clinical support.

Thirty patients were admitted due to clinical deterioration and were considered terminally ill. This group showed an expected mortality rate of 93,3% but a short mean stay period of 6,6 days. The mentioned data points at good clinical evaluation and determination of patients that are brought to in-hospital evaluation by their families or emergency services, pointing to importance of both family and medical education.

A high mortality rate of 53,30% was found on the group of patients who were not terminally ill but were admitted to hospital care due to clinical deterioration, reinforcing that clinical performance play an important role as a prognostic factor in clinical practice. Total pain, a widely known concept first described by Cicely Saunders,(7 8) was the second cause of hospital admission in our study. Patients presented a mortality rate of 60,80%, compatible with the concept of total pain per se, since most individuals in this scenario are expected to harbor advanced, highly aggressive, or refractory cancers.(7 8)

The mean stay of 9,4 days for this group was slightly larger when compared with the group admitted in terminality, namely 6,6 days.

Abdominal pain was the sixth cause for hospital admission (7,3% of all in-hospital stays). Mostly these had gastrointestinal cancers and 64,2% presented metastatic disease at admission. Mortality rates within this group - 35,7%. The mean stay for this group was 8,2 days. Multiple variables could play role on these findings, but intestinal occlusion and/ or sub-occlusion, hemorrhages, infections, total pain, and ascites were isolated and designed into separate groups. This supports the importance of careful pain evaluation in these patients.

Patient's outcomes during in-hospital stay: excluding patients in terminality, the mean stay comparison between those who showed clinical improvement with hospital discharge and those who died was surprisingly similar - 15,2 versus 13,2 days, respectively. Of important notice is the social and economic burden involved in the care of these patients. Our practice scenario is mostly composed of low-income patients who depend solely on the public health system for all health care, not mentioning that a substantial fraction of them come from nearby smaller cities that very often have poor infrastructure to assist their clinical demands in loco.

Adding to this is the lack of hospices and out-ofhospital care facilities, which directly increases the demand of hospitalized care even in cases when outof-hospital care would be better indicated.

Therefore, our admissions frequently involve socialeconomic aspects, adding more complexity to the care of these patients.

Analysis of this data, namely the mean in-hospital stay comparison between patients who were discharged and those who died, leads us to the conclusion that in-hospital care is a keystone for the care of patients in terminality or with uncontrolled symptoms.

Summarizing, the main goal of such care should focus on the well-being of these patients and their families, and not only clinical outcomes, as show inTable 3.

This is an observational prospective study evaluating a highly complex multivariable scenario and therefore has its own limitations. The dynamic physiopathology involved in malignant neoplasia, not mentioning its heterogenicity, brings a great challenge for the stratification and statistical analysis of the data. Looking further, we also have important social, cultural, and economic factors that also implicate in patient's prognosis since they influence access to health care, medicines, hygiene, family relations, and ultimately patient's well-being.

Finally, the developed “real-life” picture of our oncology infirmary does not necessarily represent the actual scenario in other Brazilian regions, each with its own set of variables, which can widely vary especially in a continent-sized country as ours.

Despite these limitations, we are hopeful that these data can bring a better understanding of the demands involved in the care of cancer patients that need hospitalization in Brazil. Our next step is to establish an instrument for directions of needs for available palliative care in oncology beds, medications, multidisciplinary staff, and development of more efficient hospice care based on our own local reality.