Hospital System Data Review Summary

This summary is created based on data from hospital_system_data.xlsx. All patient data are de-identified.

Patient
- Overview: 4,732 unique patients, with age range from 19 to 999.
- Issue: 4 patients have invalid age value 999. They have visit records across three hospitals and four doctors. It is possible that 999 is the default value for missing data, or entered by user incorrectly. Need to further review data sources, hospital records, and patient profile in EMR to identify the cause, correct data, and improve future data quality.
Visit
- Overview: 4 hospitals, 4 doctors, 4 types of diseases, visit dates cover 2015/01 - 2016/12 time frame.
- Issue: Doctor name “Jeb Alrdin” looks like a typo for “Jeb Aldrin”. In this analysis, this is corrected as “Jeb Aldrin”.

The below graph shows the number of patient visits per months per hospital. Falcon Clinic has the highest patient flow (~300 pv/month) and Kepler Medical Center has the lowest patient flow (~60 pv/month). Both of them have fairly steady patient flow acorss two years (2015 and 2016).
During 2015/01 to 2016/05, Hohmann General and St. Manley Hospital have similar patient flow volume (~125 pv/month). Since 2016/05, St. Manley Hospital has an increase trend in patient flow and it reaches 250 pv/month by 2016/12. On the contrary, Hohmann General dropped from 120 pv in 2016/05 to 3 in 2016/06, and then stopped provide services.
Of the 121 patients served by Hohmann General in 2016/05-2016/06, 37 later receive service from the other three hospitals, 84 patients stopped treatment. The change in Hohmann General’s patient flow during 2016/05 can indicate the hospital’s operation transition, management issue, or doctor loss. Need further clarify the exact cause.
We can further study in-patient vs. out-patient comparison, or comparisons across departments if given more data.

The below graph shows the number of patients treated per disease per hospital.
All four hospitals offer services in all four type of diseases.
The proportion of diseases is very similar among all four hosipitals: on average, 14.92% patients treated have Gravioli, 36.67% patients treated have Materialitis, 38.45% of patients treated have Mystery Goo, and 9.95% of patients treated have Spectrolaria.
Among all 4 hospitals, the most common disease (highest patient volume) is Mystery Goo, followed by Materialitis, Gravioli, and Spectrolaria.
Falcon Clinic has the largest number of patient treated across 2015-2016, followed by St. Manley Hospital, Holmann General, and Kepler Medical Center.

Patient age median are the same among four hospitals: 40
Patient average age per disease type: Mystery Goo has the highest patient age mean (43.0), following by Spectrolaria (41.0), and Materialitis (38.5) and Gravioli (38.1).

disease	median(age)	mean(age)	sd(age)
Gravioli	38	38.11528	6.658696
Materialitis	38	38.56046	5.849847
Mystery Goo	43	43.00400	5.227049
Spectrolaria	40	41.07536	7.118785

Each doctor specializes in one particular type of disease:
Val Armstrong specializes in Mystery Goo
Jeb Aldrin specializes in Materialitis
Bob Collins specializes in Gravioli
Gil Chavez specializes in Spectrolaria
Each doctor practices at all four hospitals, with the most services provided at each hospital ranking from high to low: Falcon, St. Manley, Hohmann, and Kepler.

For all four doctors, the monthly patient visit number is associated with the hospital size. During 2016/05 - 2016/06, all four doctors stop providing service at Hohmann General. During the same months, the monthly patient visit number increases for all four doctors at St. Manley Hospital. Jeb and Val have the most increases, which indicates that there might be an increasing volume of Materialitis and Mystery Goo patients. There’s also a minor increase with Bob and Gil’s patient visit number, which indicates an increase in Gravioli and Spectrolaria patients.

The below table shows for each disease each patient has, on average how many different hospital they need to visit and how many visits they need to make for treatment.
Mystery Goo patient visit has an average of 1.61 hospitals with average of 2.14 visits, while Spectrolaria patients visit an average of 1.11 hospitals with average of 1.19 visits. It looks like the more prevalance the disease is, the more hospital visits is needed. It is possible that Mystery Goo and Materialitis are epidemic disease and they have a higher possibility of infection. It could also be that these two type of diseases require more treatment.
We can further study patient treatment cycle for each type of disease. The average hospital visit number greater than 1 might due to patient diagnosed with same type of disease multiple times.