diff --git a/util/opentelemetry-util-genai/tests/test_handler_metrics.py b/util/opentelemetry-util-genai/tests/test_handler_metrics.py
index 6db12f0cd2..b45383abfe 100644
--- a/util/opentelemetry-util-genai/tests/test_handler_metrics.py
+++ b/util/opentelemetry-util-genai/tests/test_handler_metrics.py
@@ -1,38 +1,22 @@
 from __future__ import annotations
 
-from typing import Any, Dict, List, Tuple
-from unittest import TestCase
+from typing import Any, Dict, List
 from unittest.mock import patch
 
-from opentelemetry.sdk.metrics import MeterProvider
-from opentelemetry.sdk.metrics.export import InMemoryMetricReader
-from opentelemetry.sdk.trace import TracerProvider
-from opentelemetry.sdk.trace.export import SimpleSpanProcessor
-from opentelemetry.sdk.trace.export.in_memory_span_exporter import (
-    InMemorySpanExporter,
-)
 from opentelemetry.semconv._incubating.attributes import (
     gen_ai_attributes as GenAI,
 )
 from opentelemetry.semconv.schemas import Schemas
+from opentelemetry.test.test_base import TestBase
 from opentelemetry.util.genai.handler import TelemetryHandler
 from opentelemetry.util.genai.types import Error, LLMInvocation
 
 _DEFAULT_SCHEMA_URL = Schemas.V1_37_0.value
 
+SCOPE = "opentelemetry.util.genai.handler"
 
-class TelemetryHandlerMetricsTest(TestCase):
-    def setUp(self) -> None:
-        self.metric_reader = InMemoryMetricReader()
-        self.meter_provider = MeterProvider(
-            metric_readers=[self.metric_reader]
-        )
-        self.span_exporter = InMemorySpanExporter()
-        self.tracer_provider = TracerProvider()
-        self.tracer_provider.add_span_processor(
-            SimpleSpanProcessor(self.span_exporter)
-        )
 
+class TelemetryHandlerMetricsTest(TestBase):
     def test_stop_llm_records_duration_and_tokens(self) -> None:
         handler = TelemetryHandler(
             tracer_provider=self.tracer_provider,
@@ -52,10 +36,8 @@ def test_stop_llm_records_duration_and_tokens(self) -> None:
         ):
             handler.stop_llm(invocation)
 
-        metrics, resource_metrics = self._harvest_metrics()
-        self._assert_metric_scope_schema_urls(
-            resource_metrics, _DEFAULT_SCHEMA_URL
-        )
+        self._assert_metric_scope_schema_urls(_DEFAULT_SCHEMA_URL)
+        metrics = self._harvest_metrics()
         self.assertIn("gen_ai.client.operation.duration", metrics)
         duration_points = metrics["gen_ai.client.operation.duration"]
         self.assertEqual(len(duration_points), 1)
@@ -110,10 +92,8 @@ def test_stop_llm_records_duration_and_tokens_with_additional_attributes(
         invocation.attributes = {"should not be on metrics": "value"}
         handler.stop_llm(invocation)
 
-        metrics, resource_metrics = self._harvest_metrics()
-        self._assert_metric_scope_schema_urls(
-            resource_metrics, _DEFAULT_SCHEMA_URL
-        )
+        self._assert_metric_scope_schema_urls(_DEFAULT_SCHEMA_URL)
+        metrics = self._harvest_metrics()
         self.assertIn("gen_ai.client.operation.duration", metrics)
         duration_points = metrics["gen_ai.client.operation.duration"]
         self.assertIn("gen_ai.client.token.usage", metrics)
@@ -148,10 +128,8 @@ def test_fail_llm_records_error_and_available_tokens(self) -> None:
         ):
             handler.fail_llm(invocation, error)
 
-        metrics, resource_metrics = self._harvest_metrics()
-        self._assert_metric_scope_schema_urls(
-            resource_metrics, _DEFAULT_SCHEMA_URL
-        )
+        self._assert_metric_scope_schema_urls(_DEFAULT_SCHEMA_URL)
+        metrics = self._harvest_metrics()
         self.assertIn("gen_ai.client.operation.duration", metrics)
         duration_points = metrics["gen_ai.client.operation.duration"]
         self.assertEqual(len(duration_points), 1)
@@ -177,35 +155,29 @@ def test_fail_llm_records_error_and_available_tokens(self) -> None:
 
     def _harvest_metrics(
         self,
-    ) -> Tuple[Dict[str, List[Any]], List[Any]]:
+    ) -> Dict[str, List[Any]]:
         """Returns (metrics_by_name, resource_metrics).
 
         metrics_by_name maps metric name to list of data points.
         resource_metrics is the raw ResourceMetrics list for scope-level
         assertions (e.g. schema_url).
         """
-        try:
-            self.meter_provider.force_flush()
-        except Exception:  # pylint: disable=broad-except
-            assert False, "force_flush raised an exception"
-        self.metric_reader.collect()
+        metrics = self.get_sorted_metrics(SCOPE)
         metrics_by_name: Dict[str, List[Any]] = {}
-        metrics_data = self.metric_reader.get_metrics_data()
-        resource_metrics = (
-            metrics_data and metrics_data.resource_metrics
-        ) or []
-        for resource_metric in resource_metrics:
-            for scope_metric in resource_metric.scope_metrics or []:
-                for metric in scope_metric.metrics or []:
-                    points = metric.data.data_points or []
-                    metrics_by_name.setdefault(metric.name, []).extend(points)
-        return metrics_by_name, resource_metrics
+        for metric in metrics or []:
+            points = metric.data.data_points or []
+            metrics_by_name.setdefault(metric.name, []).extend(points)
+        return metrics_by_name
 
     def _assert_metric_scope_schema_urls(
-        self, resource_metrics: List[Any], expected_schema_url: str
+        self, expected_schema_url: str
     ) -> None:
-        for resource_metric in resource_metrics:
+        for (
+            resource_metric
+        ) in self.memory_metrics_reader.get_metrics_data().resource_metrics:
             for scope_metric in resource_metric.scope_metrics:
+                if scope_metric.scope.name != SCOPE:
+                    continue
                 self.assertEqual(
                     scope_metric.scope.schema_url, expected_schema_url
                 )