Choosing the right test to use..

Hey, I am a post-graduate research with a very basic understanding of statistics and had the following queries:

1. Testing significance of difference
Disease 1 (CMD): [N=284] mean number of contacts with patient: 16.2
Disease 2 (SMD): [N=157] mean number of contacts with patient: 28.2
Duration: 2 years.

Q:What test of significance can be used to show that the higher number of contacts in SMD is statistically significant.

2. Forecasting or Regression ?
New patients registered per year, as per an old system.
2015 - 78
2016 - 52
2017 - 30
New registrations, as per new system:
2018 - 42

Q: Is there a way to say that the number of new registration is following the trend set by the old system or if it is deviating then by how much. Using forecasting in SPSS gives an very wide window -60 to 43, which does not seem significant.

3. Calculating the correct ratio:
Number of patients continuing treatment in Disease 1
Old System [N=126]: 54%
New System [N=42]: 59%

Q: What would be the appropriate test to use to assess significant difference between the two percentages.

Any help would be much obliged.