**22. Conclusion**

Over the last two decades, prognosis has improved significantly in CF and this has led to the need to better predict outcomes. Subsequently, many studies have looked at all aspects of the disease and identified modifiable risks, as well as identified biomarkers to follow disease progression. This increased knowledge of the condition has allowed improved treatment and more phenotypically specific therapy to be developed. Many different clinical features and markers predict outcome and this is a complex area, and a multifaceted approach to risk stratifying patients is needed. However, what must be not overlooked as the survival increases in CF are the extrapulmonary manifestations and the significant psychosocial issues associated with chronic illness. It is important to note that up to 22% of CF patients demonstrate symptoms of depression, 10% report anxiety symptoms, and up to 5% report suicidal thoughts [146]. The incidence of depression and anxiety is 2–3 times higher than age matched community control individuals [147]. This is highly important and there is an association between depression and lower FEV1 [146]. Also, patients with well-preserved spirometry who have depression demonstrate more significant decline in lung function than those with no depressive symp‐ toms [148]. Depression is associated with negative medication beliefs leading to lower medication adherence, which ultimately may lead to poorer prognosis, hence it is important to evaluate patients holistically beyond the medical complications of CF when assessing prognosis [149]. Other important factors that are easily overlooked are socio-economic issues that impact upon survival, including lower household income and socio-economic status [150-152], larger family size with more than one person with CF [153], and exposure to cigarette smoke [154] all predicting a poor outcome. These risk factors are sometimes more difficult to quantify and hence are often overlooked. They are also difficult to measure as part of composite prediction tools, but they must be taken into account when assessing the prognostic indices of each individual with CF.

As survival continues to improve, the use of composite scoring systems and multi-disciplinary approach to improving care is essential in CF. Many clinical prediction tools exist for CF and some are more useful in clinical practice than others, however, these are underutilised in general and should be employed in more clinical trials, as well as routine care.

In summary, the prognosis of CF has improved significantly over the last two decades and may improve further as newer medications are developed. The need to accurately predict prognosis is essential as the decision for lung transplantation or to aggressively treat certain aspects of disease may be more tailored and appropriate per individual, this improving survival further. The most sensitive way of predicating prognosis currently remains a multi‐ faceted approach, including several markers of disease and the use of all factors and a composite clinical prediction tool is suggested to stratify patient risk.
