You can employ Time Series Analysis in SPSS.

What you need here is the Mann-Kendall test. This is a very simple test and you can implement it in any software.

Just compute the Kendall rank correlation of hosp. per month vs years.

Then see if this correlation is significant.

More info here: http://www.stats.uwo.ca/faculty/mcleod/2003/DBeirness/MannKendall.pdf

Thank you very much Andrey, I will look for it and try applying it.

Regards,

Aamir

I would suggest using R for modelling because it is more flexible and has more methods implemented in it.

If you use R, you can compute the Mann-Kendall test using only one line (for your data on the graph):

MannKendall(c(15,15,15,15,11,12,9,8,7,7,7,7,7,8,8))

Note that you need to install the Kendall package first.

The results for your data on the graph is:

tau = -0.673, 2-sided pvalue =0.0012178

The result suggests that there is a strong statistical evidence (p-value

Aamir, hi. Do you have information about the population context that provides the people that entered treatment? Is it a nation, a city, a region, or a particular community?

Other questions are: What is the main variable that you want to analyze in your time study? Is it fraction of cured people for each year, or of deseased people for each year? Is the  people that was treated per month an average of several years, or does it refer only to one year?

Or is your may question completely different from those I mentioned?

OK, best wishes. Emilio

Thank you very much Andrey. I'm using SPSS and don't have R so it won't be possible. 

Could you please connect with me via email or skype?

Dear Emilio,

I will be answer your questions as:

1. The study is conducted in a Tertiary care hospital in a district of around 0.2 million population

2. I want to see the trend of cases that were attended in the hospital for poisoning over the five period time i.e. 2011-2015.

3. I want to see if there is improvement (increase) in the number of cases cured over time.

Regards,

Aamir

What's your e-mail?

[email protected] 

Dear Aamir,

You can also try TREND, a free software specifically designed to facilitate statistical testing for trend, change and randomness in time series data. TREND has 12 statistical tests including Mann-Kendall test. Visit http://www.toolkit.net.au/trend.

Regards, Dweependra

Cool soft, Dweependra !

Thank you Dweependra I will surely follow your suggestion!

I have done a trend analysis using MS Excel already and it sounds fine.

@ Andrey I was able to find the option for trend analysis in spss also and did it there too but the graphics in spss are not that good as that of MS Excel. Will soon share it with you here.

Aamir

You can employ Time Series Analysis in SPSS.

Zainudin is right