diff --git a/pyproject.toml b/pyproject.toml
index 919c1aa..c84b873 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -16,7 +16,7 @@ classifiers = [
 dynamic = ["version"]
 requires-python = ">=3.9"
 dependencies = [
-    'nested-pandas==0.1.2',
+    'nested-pandas==0.1.3',
     'numpy',
     'dask>=2024.3.0',
     'dask[distributed]>=2024.3.0',
diff --git a/src/nested_dask/utils/utils.py b/src/nested_dask/utils/utils.py
index e6c250d..163f97b 100644
--- a/src/nested_dask/utils/utils.py
+++ b/src/nested_dask/utils/utils.py
@@ -35,8 +35,6 @@ def count_nested(df, nested, by=None, join=True) -> NestedFrame:
     if by is not None:
         # will have one column per unique value of the specified column
         # requires some computation to determine these values
-        # TODO: Requires modification of nested-pandas to always produce
-        # sorted output columns for meta
         by_cols = sorted(df[nested].nest.to_flat()[by].unique())
         out_cols = [f"n_{nested}_{col}" for col in by_cols]
     else: